Value Stream Mapping for Lean Manufacturing
- Arpit Shah
- Nov 23, 2023
- 72 min read
Updated: 2 days ago
1. INTRODUCTION
Regardless of whether you end up choosing to prepare Value Stream Maps, just coming to terms with the apparatus of this impressive Lean Manufacturing tool can be a rewarding experience.
Lean philosophy recommends organizations to pursue operational excellence by eliminating Waste from their operations and by focusing on Continuous Improvement with the objective to get the right products, to the right place, at the right time, in the right quantity, and in the right condition. To put it simply, in a world of limited resources and numerous constraints, an organization must strive to maximize its efficiency.
While Value represents what a customer is prepared to pay, a Value Stream represents the flow of Material and Information pertaining to a Product Family - from the supply of Raw Material inputs till the dispatch of Finished Goods to the customer.
Mapping the Value Stream as-is (Current State) results in the development of a macro-level diagram depicting the span of an organization's operations through which one is sensitized about how Value is created and where Waste lies in the system so that one can contemplate ways to improve the former and eliminate the latter. This introspection, which also takes into consideration aspects such as anticipated demand, customer preferences and the resources and technology needed to cater to it, culminates into a to-be design (Future State) of the Value Stream. This ideal/improved rendition of the Material and Information Flow forms the basis of, and lends direction to, an organization's manufacturing and overall business strategy.
HYPERLINKS TO SECTIONS
1. Introduction
2.3 Bullwhip Effect
4.1 Case Background
4.5.1 The Bottle Neck Analogy
1. Continued
Alongside how to perform Value Stream Mapping (which I will be covering in-depth), it is also important know where to use it too.
Assume that you are tasked with mapping the Value Stream of Mapro - a renowned Indian manufacturer of fruit-based confections such as candies, syrups, jams and mixers.
Would you map its entire factory i.e. the manufacturing operations of all products?
Would you map an individual process such as Boiling or Mixing?
Would you map the entire supply chain i.e. all operations and activities from farmer to consumer?
The answer to all of the above is No. Instead, Value Stream Mapping is carried out for a Product Family and its scope extends to the manufacturer's span of direct influence.
A Product Family is those group of products that pass through similar processing steps over common equipment, particularly in the manufacturing processes situated downstream. At Mapro, three distinct Product Families can be observed-
Note: All images in this post are zoomable and downloadable
Implementing the operational improvements established from the Value Stream Mapping study in the timeline envisaged is only possible for those activities which the manufacturer can exert sufficient control over. Hence, a manufacturer's span of direct influence typically extends from its immediate supplier at one end to its immediate customer on the other.
For example, Mapro can influence its Raw Material quality with the Sugar Mill (supplier) much more readily than it can influence the Farmers (supplier of supplier) to cultivate a different breed of Sugarcane. Similarly, it can influence the buying behaviour of Wholesalers (customer) by making policy-changes such as revising the MOQ much more readily than it can influence the Retailers (customer of customer) into altering their buying behaviour. And goes without saying, Mapro can certainly add a new packing machine at its own plant if it wishes to produce a different configuration of candy pouches. Hence, it is practical that Mapro extends the scope of Value Stream Mapping from its supplier upstream to its customer downstream.
2. OPERATIONS AND SUPPLY CHAIN ASPECTS TO KNOW
2.1 PRODUCTION CONFIGURATION
A Product Family's demand characteristics, among other factors, influences a manufacturer to adopt a particular type of Production Configuration. For example, consumers expect Mapro candies to be available at a nearby retail outlet whenever they crave a sweet treat and hence, by virtue of being a fast-moving consumer product, the entire candy distribution network has to be set up to be agile. Distributors, Wholesalers and Retailers would all like their Customer Lead Time i.e. the time it takes to receive the consignment after placing a Purchase Order to their respective supplier, to be short. Even Mapro, which receives Purchase Orders from several Wholesalers every day, strives to fulfill its customer demand in quick-time by dispatching candy consignments on the same day itself.
The demand for candies is largely stable, barring a significant surge during festive season. Mapro predicts the demand for its candies using Forecasting software which utilizes inputs such as historical orders, current market trends, competitor performance and point-of-sale data. The derived forecast is subsequently used to prepare the Production Plan and Schedule i.e. how many candy pouches to produce and when, in order to fulfill the customer demand on-Time and in-Full (OTIF).
This type of Production Configuration is called Make-to-Stock or Build-to-Stock with Forecast (refer Figure 5). In this approach, a manufacturer mass produces Finished Goods preempting Customer Demand. Waiting for confirmed customer orders to arrive before beginning to produce the candies would increase the Customer Lead Time of the Wholesaler, which would be highly unfavorable given the nature of the industry.
Contrastingly, think of Fenesta (a leading Windows and Doors manufacturer in India), where once a customer expresses interest to buy Windows, the Presales team visits the site and obtains the measurements which are subsequently fed into a Bill of Material software to determine the exact quantity of material needed (and its cost) to manufacture it. A quotation is subsequently prepared and submitted to the customer. Only after order confirmation (and an advance payment) does Fenesta initiate manufacturing the Windows. Since the primary raw materials are few and standardized (eg. uPVC granules, Aluminium, Fittings), Fenesta keeps ample Inventory of these at its warehouse, readily deliverable to the factory as and when required.
As all the processes, barring production design and raw material procurement, are initiated only after receiving order confirmation from the customer, this type of Production Configuration is known as Make-to-Order. The customer is prepared to wait too i.e. has lenient Lead Time preferences, recognizing that it would take some time for the personalized furniture to be manufactured, delivered and installed at his premises.
This Business Strategy of delaying investments into certain manufacturing activities till confirmed orders arrive is known as Postponement and the degree of Postponement utilized is reflected in the Production Configuration that an organization chooses to adopt. Push vs. Pull considerations i.e. whether to reap the benefits from Economies of scale by manufacturing in large batches or whether to reduce the exposure to Demand volatility by manufacturing small quantities, play an important role in determining the suitable Production Configuration. Each type comes with its own benefits and risks - a manufacturer must carefully evaluate the trade-offs involved.
Transitioning to a different Production Configuration is usually gradual i.e. one step on either side, at-a-time. Exceptions such as Dell, which made a radical and successful switch to a Make-to-Order strategy with a Just-in-Time Inventory system in an industry laden with Make-to-Forecast manufacturers, are a rarity.
2.2 THE 7 WASTES OF LEAN PRODUCTION
Eliminating Waste (of the Type 2 variant in the Japanese concept of Muda) from the Value Stream is fundamental to Lean Manufacturing. Waste comprises of seven, often interrelated, constituents-
While observing the presence of all the types of Waste across the Value Stream is critical, spotting and eliminating its source is fundamental - which typically tends to be Overproduction - widely considered to be the deadliest type as it spurs the generation of the other types of Waste, be it excess Inventory, unnecessary Transportation, increase in Waiting Time or more Defects. Producing more than what is needed is a malaise which often stems from unreliable Forecasts or from the misguided appeal of producing in large batches in order to reap the benefits from Economies of Scale.
2.3 BULLWHIP EFFECT
The Bullwhip effect highlights how small fluctuations in consumer demand can trigger large fluctuations in the ordering behaviour upstream in the supply chain network, eventually culminating into significant Overproduction by the manufacturer. The infographic below illustrates this phenomena using fictitious demand data from the supply chain network of Mapro Syrups-
To summarize the depiction-
While the daily demand for Mapro Syrups from the consumers that Retailer 1 serves averaged 30 bottles for the week (150 bottles sold across 5 working days), Retailer 1 placed an order of 10 cases (200 bottles) for the upcoming week to Wholesaler 1 i.e. an overorder of 50 bottles vis-à-vis historical demand
In response to the weekly demand totalling 1000 bottles from the five retailers it serves, Wholesaler 1 placed an order of 12 crates (2400 bottles) for the upcoming fortnight to Mapro i.e. an overorder of 200 bottles per week vis-à-vis historical demand
In response to the fortnightly demand totalling 12000 bottles from the five wholesalers it serves, Mapro produced a truckload of syrups (30000 bottles) for the next month i.e. an overproduction of 3000 bottles per week vis-à-vis historical demand
The root cause which triggered this spate of overordering across the supply chain network and which culminated in overproduction by the manufacturer was the daily fluctuations in consumer demand for Mapro Syrups. Also notice that the demand-to-supply mismatch amplified as one went upstream in the supply chain network and the manufacturer had the severest reaction of them all - it overproduced 3000 bottles per week vis-à-vis historical demand from its wholesalers. But this is not the true picture - in reality, Mapro has overproduced 3750 bottles per week vis-à-vis historical consumer demand (Production of 30000 bottles in a month − Sale of 600 bottles in a month per retailer x 5 retailers x 5 wholesalers ÷ 4 weeks).
Assuming that the daily fluctuations persist but the total consumer demand remains the same in the upcoming month, we would likely witness a reversal in the ordering behaviour - a spate of underordering across the supply chain network, one which amplifies upstream and eventually culminates in underproduction by the manufacturer. This is because each node in this three-echelon supply chain network (Retailer, Wholesaler, Manufacturer) would realize that it has excess Inventory of Mapro Syrups from the previous month to sell and hence, would order/produce less vis-à-vis its historical customer demand. And just as it routinely occurs in the marketplace, deep-discounting may be resorted to in order to clearout the excess Inventory and free up the Working Capital - Syrup being a perishable commodity would only serve to fuel this urgency. Bullwhip effect wreaks havoc far and wide!
But wouldn't Retailer 1 been better off placing an order of 8 cases (160 bottles) to the wholesaler instead of the 10 (200 bottles) it did in response to the consumer demand of 150 bottles?
In hindsight, Yes. However, here are some of the reasons which would have influenced Retailer 1 into overordering-
Fluctuations in consumer demand throughout the week, with as high as 60 bottles being sold in a day, influenced the retailer to overcompensate for the upcoming week. This is a natural tendency as it is preferable to have excess Inventory rather than turn customers away due to stock-out
The wholesaler's Minimum Order Quantity (MOQ) may have been 10 cases, leaving the retailer with no choice but to order this quantity
The retailer may have wilfully ordered this quantity to qualify for rate discount and/or incentives
The Syrup consignment arrives once-a-week and ordering extra would serve as a safety net against possible logistical delays
Similar reasons would induce overordering upstream in the supply chain network and the mismatch from consumer demand would only widen. To avoid falling into the trap of overproduction due to Bullwhip effect, a manufacturer must strive to have rapid visibility of point-of-sale data i.e. records of consumer purchases at retail-level, which would help to determine the actual demand and produce to it. It should also share this information across the distribution network so as to mitigate overordering.
3. VALUE STREAM MAPPING SYMBOLOGY
Those who wish to know about the origins of Value Stream Mapping may read this resource. The infographic below depicts commonly-used icons used in preparing a Value Stream Map, with a short description for each and categorized based on the type of activity they represent-
The file containing this depiction (MS Visio needed to edit/modify the icons) can be downloaded here.
4. VALUE STREAM MAPPING CASE - ACME STAMPING COMPANY
Credits - Learning to See: Value Stream Mapping to Add Value and Eliminate Muda book by Rother and Shook; Lean Supply Chain MBA Course by Professor Arvind Subramanyam
4.1 CASE BACKGROUND
The business entity in concern - Acme Stamping Company - manufactures automotive components. It wants to map the Value Stream of the Stamped Steel Bracket Subassembly Product Family.
This Product Family consists of two products that fit underneath the instrument panel-
1. Bracket for Left-Hand Drive (LH) vehicle
2. Bracket for Right-Hand Drive (RH) vehicle
These products are sold to one customer - State Street Assembly, an automobile manufacturer.
4.2 DRAWING THE CURRENT STATE VALUE STREAM MAP
Assume that you are a manager at Acme and are tasked with driving this Value Stream Mapping study. You have full access and authority to obtain information pertaining to the organization's manufacturing operations straddling multiple departments. Grabbing a sheet of A3 paper, stationery and a timer, you take a tour of the manufacturing unit to audit the processes that operate on Steel Brackets. Your initial objective is to obtain data pertaining to the Material Flow.
You begin by visiting the process that lies farthest downstream in the Steel Brackets Value Stream - Shipping department. Upon acquiring data pertaining to the customer-facing operations here, you will work your way backwards i.e. upstream, collecting manufacturing operations-related data along the way. You will complete your audit by gathering supplier-facing operations data at the Stamping process situated farthest upstream; it receives the only Raw Material that goes into the manufacturing of Steel Brackets - Steel Coils - from a supplier called Michigan Steel Co.
From the information gathered at the Shipping department, you draw a Customer/Process box on the far right of the A3 paper and label it with the name of the customer. Below it, you add a Data box and fill it with key metrics pertaining to the customer. State Street demands 18400 Brackets from Acme each month - 12000 LH and 6400 RH (~ 2:1).
The Brackets are sold in denomination of Trays - this is the Pack Size. A Tray comprises 20 Finished Brackets, either exclusively LH or RH.
Upto 10 returnable Trays can be loaded onto a Pallet, which is the container type used for dispatch. However, you choose not to make a mention of this information on the Map.
Note: Value Stream Mapping entails macro/high-level charting of the Material and Information Flow. All additional information and details can be preserved separately for reference.
State Street performs its automobile manufacturing operations across two Production Shifts in a day and has worked out a supply arrangement with Acme who ships a consignment of Finished Steel Brackets to its factory on a daily-basis.
Next, you proceed to acquire information pertaining to the processes involved in the manufacturing of Steel Brackets. Each discrete manufacturing process i.e. those set of activities where material flows uninterruptibly between, needs to be mapped separately.
Figure 12 below depicts the entire manufacturing sequence of the Steel Brackets Value Stream with the requisite Process Boxes and Data boxes. The five discrete manufacturing processes that you've identified are - Stamping, Spot Weld 1, Spot Weld 2, Assembly 1 and Assembly 2.
Note: A Value Stream Map does not display the physical layout of the factory, it just displays the sequence of operations.
What do the red triangles and the striped arrows between the Process Boxes represent?
The triangles signal that Inventory is stagnating/accumulating before each of the five manufacturing processes prior to being operated upon - this can also be indirectly inferred from the fact that if material is flowing uninterruptibly within each of the discrete manufacturing processes i.e. input material enters, gets operated upon, and exits as output without pausing anywhere in-between, then surely it must stagnate outside the realm of these processes. What is actually important is the label underneath the triangle which depicts the quantity of stagnating Inventory and/or the duration of stagnation observed. For example, during your audit you had observed that 4600 units of stamped LH Brackets and 2400 units of stamped RH Brackets had accumulated before the Spot Weld 1 process.
The Stamping process is where Steel Coils (Raw Material) gets pressed into RH and LH Steel Brackets. Hence, downstream of Stamping you've chosen to depict stagnating Inventory in quantity terms. Prior to Stamping, as the material is not yet transformed into the Product Family, you chose to depict the stagnant Inventory in duration terms (5 days-worth of Steel Coils). Note: the cost of stagnant/excess Inventory is not mapped.
The striped arrows signify Batch-Push Material Flow i.e the manufacturing process upstream produces output for its downstream counterpart in large batches as per a Production Schedule issued by the Production department. Basically, Push signifies that the thrust of Material Flow is exerted by the process producing the material and not by the one consuming it.
The number of manpower operating a discrete manufacturing process is also depicted within its Process box. As evident from Figure 12, all processes pertaining to Steel Brackets are being run by one Operator each.
The Data box underneath the Process box of a discrete manufacturing process captures vital parameters pertaining to the Material Flow in about 3-5 rows. You have chosen to depict-
Cycle Time (C/T): This is the rate at which a manufacturing process releases a single unit of output. For example, a C/T of 1 second at the Stamping process indicates that the Stamping Press equipment churns out Stamped Brackets at the rate of one-per-second when operational
Changeover Time (C/O): This is the time it takes for a process to set up into manufacturing another product from the Product Family i.e. from LH to RH Brackets or vice versa in Acme's case. Changeover Time includes the time taken to set up the equipment, for input material to become available, for the operating manpower to be ready, and so on
EPE (every part every...) or Production Batch Size: This is the time interval within which a product from the Product Family is manufactured by the process. For example, an EPE of 2 Weeks at Acme's Stamping process indicates that the Stamping Press equipment changes over into manufacturing a Stamped LH Bracket or a Stamped RH Bracket once-every-two-weeks
Uptime: This denotes the on-demand availability of the equipment or the overall reliability of the process in general. For example, an 85% Uptime at the Stamping process indicates that the Stamping Press equipment is out-of-service i.e. is facing downtime for 15% of the available working time. This data is typically documented by the equipment operator or can be pulled from the equipment log - obtaining this data through observation is not the right approach
(Production) Shifts: This is the number of man-days of manufacturing that occurs within a single workday. The Steel Brackets Product Family is manufactured in 2 Shifts at Acme i.e. 2 man-days of 8 hours each in a workday
Working Time Available per Shift: Self-explanatory. Expressed in seconds so as to avoid any confusion arising from a fraction/decimal-based expression. At Acme, 20 minutes is reserved for Lunch break in an 8 hour-Shift. Thus, the available working time per Shift is 27600 seconds
While these process parameters are sufficient to capture the key aspects pertaining to the Steel Brackets Product Family at Acme, one can also choose to capture other parameters such as-
Production Lead Time: This is the length of time it takes for an input to be processed into output (this parameter is separately depicted in the Timeline section of a Value Stream Map)
Rework rate: This is the proportion of output that needs to be reprocessed/repaired/replaced vis-à-vis total output
Scrap rate: this is the proportion of discardable output being generated vis-à-vis total output
At the bottom of the map lies the Timeline section which depicts the Production Lead Time at each distinct stage of the Value Stream. The elevated portion of the Timeline corresponds to the space between two Process boxes and its label denotes the length of time a unit of Inventory lies idle/waiting at this juncture - derived by dividing the total quantity of Inventory stagnating here by the average customer demand per day. For example, 7000 Stamped Brackets have been observed as stagnating between Stamping and Spot Weld 1 processes. Given that State Street's average daily demand for Finished Brackets is 920 units (18400 units in a month ÷ 20 workdays), this translates to an outside-process Production Lead Time of 7.6 days (7000 ÷ 920) as depicted in Figure 12.
The depressed portion of the Timeline corresponds to the space within a Process box and its label denotes the length of time it takes for a unit of input to be processed into output i.e. the within-process Production Lead Time or simply, the Processing Time. In Acme's case, the Processing Time is equal to the Cycle Time for all the manufacturing processes as the inputs are processed individually by the operator/equipment everywhere in the Value Stream.
The sum of all the outside-process and within-process Production Lead Times (i.e. Total Production Lead Time) as well as just the sum of all the within-process Production Lead Times (i.e. Total Processing Time) is depicted at the end of the Timeline section. This bifurcation helps in knowing what proportion of the overall Production Lead Time does a unit of material actually spend accreting Value.
Note: The 23.6 days Total Production Lead Time depicted in map in Figure 12 is not just the sum of all the outside-process Production Lead Times alone - it includes the sum of all within-process Production Lead Times, i.e. the total Processing Time, as well. The illusion occurs as the latter is a very tiny figure (188 seconds) compared to the former.
Next, you proceed to acquire information pertaining to the supply-side operations which is the only remaining Material Flow to be mapped. Michigan Steel Co. ships Steel Coil consignments twice-a-week (every Tuesday and Thursday) to the Stamping process at Acme.
Now that the mapping of Material Flow is complete, you proceed to gather data pertaining to the Information Flow. Doing so shall complete the creation of the Current State Value Stream Map.
Acme's Production Control department receives three Demand Forecasts from State Street every month - the total quantity of Finished Brackets that it estimates to procure in the next 90 days, 60 days and 30 days respectively. Actual Purchase Order is issued once every day
Production Control, a centralized Production department, plugs the Forecast and daily Order data into its Material Requirement Planning (MRP) software which generates and releases a Weekly Schedule, i.e. production target for the week, to each of the five manufacturing processes as well as a Daily Ship Schedule, i.e. shipping target for the day, to the Shipping department
The MRP software also generates and releases Acme's own Demand Forecast for Steel Coils for the upcoming six weeks to Michigan Steel Co. Actual Purchase Order is issued once every week
Communications with external stakeholders (to supplier and to customer) occur through electronic mediums such as email and fax - denoted using wiggly lines, whereas internal communications (to operators) is transmitted manually/in writing - denoted using straight lines
The Current State Value Stream Map for Acme's Steel Brackets Product Family is now complete (Figure 15 below). What insights can you draw from this representation? Where are the issues and how can they be remediated? You may want to pause to think here...
4.3 FUTURE STATE DESIGN OF A VALUE STREAM
The Future State Value Stream is designed with the intention to inculcate Lean Manufacturing. While the Current State's Material and Information Flow was mapped using the information acquired from the Plant audit, the Future State's Material and Information Flow will be developed based on the insights drawn from analyzing the Current State Map, by referring to Lean Manufacturing guidelines, and from Acme's own assessment regarding the dynamics pertaining to resources, technology and customer behaviour and preferences in the future. Several versions of the Future State Design may be iterated before settling on a particular one that is pragmatic to implement as well as is in-sync with the business strategy and goals of the organization.
Let me reveal the selected Future State Design of the Steel Brackets Value Stream at the outset itself-
4.4 UNPACKING ACME'S FUTURE STATE DESIGN
Summarizing the factual information depicted in the Future State Value Stream Map (Figure 16)-
Material Flow
A Continuous Flow workstation called Weld+Assembly shall perform both the Spot Welding as well as both the Assembling activities and it shall be managed by three operators in total
Three Supermarket Pull Systems will be deployed which shall regulate the production of Steel Coils, Stamped Brackets and Finished Brackets respectively
Four Kaizen Bursts will be scheduled: three for Weld+Assembly process (improving Uptime of equipment at Spot Weld 2, improving Changeover Time of Spot Welding operations and improving Operator utilization at the workstation) and one for the Stamping process (improving Changeover Time of Stamping operations)
Inbound Raw Material consignments would arrive once every day and outbound Finished Goods consignments will continue to be dispatched to the customer once every day
Information Flow
Production Control Department will issue Production instructions to Weld+Assembly at Pitch intervals of 20 minutes
Demand Forecasts from State Street as well as Demand Forecasts for Michigan Steel shall follow the same cadence as is prevalent currently
State Street will continue to issue Purchase Orders to Acme once every day; Acme will issue a Purchase Order to Michigan Steel Co. once every day
Pursuant to the implementation of this revised design of Material and Information Flows, Acme anticipates the Total Production Lead Time for its Steel Brackets Product Family to become 5 days (79% reduction), and the Total Processing Time to become 169 seconds (10% reduction).
4.5 WHAT DOES DESIGNING A FUTURE STATE VALUE STREAM ENTAIL?
What prompted Acme to revise its Material and Information Flows in this manner? What was the methodology used to identify Waste and introduce countermeasures? How does Acme benefit by adopting Lean Manufacturing?
While I will address these pertinent questions shortly, know that designing a Future State Value Stream is as much an expression of imagination and craft as it is an exercise in logical thinking. There are Lean Manufacturing guidelines to refer to, formed from the insights derived from previous implementations which act as an ambit within which iterations of the Future State Design can reside. Once consensus is achieved among the stakeholders for a particular iteration and is signed off by the leadership group, the implementation phase to transition to the revised system within the timeline envisaged commences.
Lean Manufacturing is an interplay of several technical concepts which warrant a deep level of understanding and I will attempt to explain it in sufficient detail. That being said, you will also need to make the effort to visualize the connection between Lean Manufacturing and the Future State Design as I feel that this linkage can neither be explained nor understood through theory alone. To assist you in this regard, as well as to prime you for what lies ahead, I've developed three analogies-
4.5.1 THE BOTTLE NECK ANALOGY
Bottleneck is a term commonly-used in Operations parlance in the context of impeding productivity. Easing the bottleneck(s) is a vital step in a bid to improve the Material Flow and to remediate the Value Stream. Let me elaborate this using...well...literally a Bottle's Neck😊.
In a regular bottle, the neck has the same girth as the mouth of the bottle - a very useful design feature as it throttles the outward flow of the liquid from the broad bottle, allowing a person to drink it or pour from it at a convenient pace. However, if the bottle's neck i.e. liquid flow rate were to be narrower than the bottle's mouth i.e. liquid consumption rate, then it would become an impediment to satisfaction.
Imagine drinking from this fancy bottle (Figure 18 below), whose neck is situated in the middle and is much narrower than the mouth, on a blazing summer day. The liquid to the right of the neck would flow into your mouth at a breezy pace and you would gulp it with glee. However, once this section empties and the liquid from the left of the neck enters your mouth, you would feel an immediate sensation of dissatisfaction - as the liquid is flowing out at a much slower pace than what you desire to consume!
Now transition to a Value Stream context - consider yourself as the consumer and the fancy bottle as being representative of manufacturing operations. Are you able to draw the parallels?🤔
If any activity throttles the rate of output below the consumer's rate of product demand, then that activity is a Bottleneck.
But what if the consumer is not inconvenienced by the reduced flow rate?
Then in that case, the Value Stream would be considered devoid of bottleneck(s). However, this may not be a reason to celebrate as the manufacturing operations may be overproducing instead! This is because if the weakest manufacturing process is producing output at a rate that is satisfactory for the customer, then the other manufacturing processes situated upstream may be producing output at a pace that is higher than what the consumer is prepared to buy. This tantamounts to overproduction as it is not restricted to producing excess of Finished Goods alone - producing excess of Work-in-Process Inventory is also considered as so (Overproduction, as I've indicated earlier, is the deadliest type of Waste). Also, while the broad bottle has its utility in that it is a repository for the precious liquid, the manufacturing operations needn't be a repository of excess Inventory. Instead, efforts must be directed towards synchronizing the rate of production to the rate of consumption across the Value Stream as much as is pragmatic to do so.
Inventory can be of three types (Raw Material, Work-in-Process and Finished Goods) and it represents the cost of capital and resources invested on aspects such as material, processing, operators, sorting, auditing, transportation, handling, storage and financing. Till the time excess Inventory is converted into Finished Goods and is purchased by the customer for which full payment is made, the investment is not realized. This impedes the Operating Cash Flow and increases the Working Capital requirements, thereby increasing the cost of Production and lowering overall Profitability. There are Opportunity Costs to consider as well - as the capital could have very well been invested on more lucrative Product Families, on some other Projects or just lie in a bank account attracting interest. While the cost of stagnant Inventory is not depicted in a Value Stream Map, its negative impact can be gauged in duration terms from the elevated portion of the Timeline section i.e. the outside-process Production Lead Time.
To summarize, using the Bottle Neck analogy, I've conveyed three important aspects of Lean Manufacturing which are also the objectives when designing a Future State Value Stream-
Knowing the rate of customer demand allows a manufacturer to identify the bottleneck(s) in the Value Stream i.e. those processes that are churning output (Cycle Time) at a slower pace than the demand rate for Finished Goods (also known as TAKT or TAKT Time which is an all-important Lean Manufacturing parameter derived by dividing the total available working time by the total demand for Finished Goods). The location of the bottleneck(s) in the Value Stream adversely impacts the pace of Production of the processes downstream, converting them into bottlenecks as well despite the fact that they may be having desired processing capacity
While the bottleneck(s) causes the manufacturer to underproduce which is undesirable, having no bottleneck(s) could indicate that one or more processes may be producing output at a rate higher than that of customer demand. This tantamounts to overproduction and it can occur anywhere in the Value Stream and not necessarily just at the Finished Goods stage. Lean philosophy despises Overproduction as it spurs the generation of Waste in its other manifestations. Instead, a manufacturer must aim to produce output at a pace which coincides with the rate of customer demand across the Value Stream as much as is pragmatic to do so
Reducing the Production Lead Time associated with stagnant Inventory would allow the manufacturer to flip the capital invested in Production into Sales faster - known as Inventory Turnover in Accounting. An improvement here would free up Operating Cash Flow, reduce Production costs and increase overall Profitability
4.5.2 THE LIBRARY CARD SYSTEM ANALOGY
The way the traditional Library Card System (Newark System of Charging and Discharging) works makes for a fine analogy to demonstrate the utility of reliable information signalling in a Lean Value Stream. Knowing the intricacies of this simple yet profound methodology will aid you in understanding the valuable Supermarket Pull System, elaborated later, with more clarity.
In the Newark Library Card System, a borrower who wishes to loan a book obtains it from the shelf and proceeds to submit it at the librarian’s desk for record-keeping along with his Borrower's Ticket. The librarian extracts the Book's Card from its Holder within the book, makes a data entry in it with the borrower's details and inserts it in the Borrower's Ticket pocket before filing it until the book is returned. Thereafter, the librarian will stamp the return date on the Due Date Sheet located on the book's flyleaf and subsequently hand over the book to the borrower. Lastly, the borrower will scribble the book's name and its due date in his Book Issue Card and present it to the librarian who will sign it as an acknowledgment before handing it back to him - when the book is returned, the librarian will scratch off this data entry and place the returned book in the designated bookshelf at the library.
What purpose does this system serve?
This nifty signalling system is of high utility to both, the librarian as well as the borrower-
It is an elaborate system of recording a transaction which allows the librarian to know how many books have been issued and how many books are lying in the bookshelves at any given point in time. This saves the need to undertake a physical audit of the library
The librarian is typically seated either at the entrance or near the bookshelves which serves as a useful surveillance spot, one that signals the visitors to get the borrowing procedure done without fail. This prevents the scope of of books going missing or records being incomplete
The borrower can refer his Book Issue Card to know how many and which books he has borrowed, and when to return them. This spares him from the need to find and audit the Due Date Sheets within each and every book that he has borrowed
In a Lean Manufacturing setup, the manufacturing processes will pull Production targets in small batches throughout the workday based on the consumption of their respective downstream counterparts instead of manufacturing large batches based on a Schedule issued to it by a Production department. This is where the acute need for a reliable Information Signalling system is felt in order to have the manufacturing processes operate at high efficiency and devoid of Waste.
A reliable information signalling mechanism would help to-
systematically transfer Production instructions from one process to another – akin to the transfer of book-related information between the borrower and the librarian in this analogy
diminish the need to conduct physical audit of the manufacturing operations - just as the librarian benefits from the Library Card System
enable a circular transaction loop with checks and balances – akin to the various components of the Library Card system. Elaborated in detail later
enhance operational efficiency – a Supermarket keeps a close tab on the consumption rate of a manufacturing process downstream which helps it to regulate production upstream - akin to how the librarian's desk serves as a vantage point to surveil visitor activity and maintain compliance
4.5.3 THE ONE DAY INTERNATIONAL (ODI) CRICKET CHASE ANALOGY
Through this analogy, several technical aspects of Lean Manufacturing, and more importantly their inter-relatedness, can be unpacked in an easy-to-grasp way (particularly if you follow the sport).
As can be observed from the Match Scorecard in Figure 20, England scored 302 runs batting first in a One Day International (ODI) format Cricket match. Australia knows that in order to chase down England's challenging score (which it managed to successfully), it would need to score at a rate marginally above a-run-a-ball (a total of 300 balls are available to chase down the target of 303).
The Australian cause would stand to benefit if their batsmen keep their rate of scoring in the vicinity of the required rate throughout the chase. If it is slower, then chasing down the target would become an uphill, and eventually an improbable task. On the contrary, scoring aggressively would carry the risk of losing precious wickets, which would also imperil the chase. Hence, timing the chase becomes vital - the batsmen would do well to align themselves to the target, while also having enough flexibility to be able to modify their approach whenever necessary based on their own as well as their opponent's strengths and weaknesses.
From a Lean Manufacturing context, think of England as the customer and the target they've set as being representative of their demand. Chasing down the target may be challenging for Australia - the manufacturer - but it is imperative that it fulfill the customer demand in order to win the match of business through increased revenue, market share and profitability.
As indicated in the Bottle Neck analogy, the manufacturer would do well to produce output at a rate that is similar to the rate of customer demand. Being aggressive and producing at a faster pace will lead to overproduction, an abominable sin, whereas being doing the opposite would cause the manufacturer to underproduce, preventing it from fulfilling the customer demand.
Would it be pragmatic for the Australian coach to issue a circular to each and every batsman at the beginning of the chase: Player 1 - score fifty runs, Player 2 - score seventy runs, and so on?
This would rarely work out as planned - professional sport mimics real life. Rather, the coach would be better served if he breaks down the target into small milestones and issues scoring instructions to only the two batsmen out at the playing crease at regular intervals throughout the chase.
Similarly, Lean philosophy recommends that rather than a Production department issuing a Schedule (a large Production target) to all the manufacturing processes, it should release instructions to manufacture a small quantity at regular intervals throughout the workday, and preferably to just one manufacturing process (typically the one nearest to the customer) which would become the Pacemaker of the Value Stream facilitating consumption-to-production synchronization for the processes upstream. Overall, Pull will be cultivated in the Value Stream as the thrust of Material Flow would be regulated by consumption and not exerted by production.
To explain using an example, Process Z, the Pacemaker of the Value Stream, must not receive a Schedule stating produce 7500 units of Product 1 in the week. Rather, it should be issued an instruction - produce 50 units of Product 1 at regular intervals throughout the workday. Process Y upstream should neither be issued a Schedule nor given small production instructions. Instead, it should attach its rate of production to the rate of consumption of Process Z downstream. This Pull-based Material Flow should replicate further upstream across the Value Stream - between X and Y, W and X, and so on as much as is pragmatic.
This Levelling of Production Volume would ensure that the manufacturing processes do not produce large batches of output which would not only prevent the generation of Waste but would also help in responding effectively to Demand Volatility, an issue that plagues most industries. The Pull-based Material Flow would also foster teamwork as the process operators would be compelled to remain zoned in towards achieving the small production targets issued to them by their downstream counterpart rather than hunting down a large target in a silo.
Being tethered would also help the process operators to realize that their efforts are directly contributing towards the organization being able to achieve its objective of fulfilling customer demand on-time. This cultivates TAKT Image across the Value Stream i.e. a shared sense of responsibility towards fulfilling the customer objective (TAKT, as I had alluded to earlier, is a miniaturized parameter that reflects the process performance target vis-à-vis customer demand). By contrast, the Push-based Material Flow approach, where a Schedule is issued to all manufacturing processes, deters the formation of TAKT Image as the process operators would become prone to working as individual entities interested in only achieving their own target however they can rather than serving a common cause in a synchronized way. The farther upstream a manufacturing process is situated, the more aloof it is susceptible to become to customer needs i.e. have a fainter TAKT Image. This triggers the creation of all types of Waste in the Value Stream, be it Overproduction, excess Inventory, increase in Waiting Time and/or more Defects.
The small target instructions released to the Pacemaker process at consistent intervals is known as Pitch in Lean Manufacturing and it is expressed in time terms. Pitch is either equal to or a derivative of a relevant production parameter, typically TAKT and a relevant consumption parameter such as Pack Size/Minimum Order Quantity/Average Order Size. This helps ensure that the target is not only a small quantity per se but also practical to conform to. In cricketing terms, this is akin to the Australian coach issuing an instruction - score 50 runs in the next 8 overs - which is both, a small quantity as well as a practical instruction for the two batsmen on the playing field. By comparison, an instruction such as - score 45 runs every 44 balls - would be difficult to conform to despite it being a small target as it is a not a simple expression in cricketing parlance, one that is difficult to ingest.
Lastly, the Australian batsmen would bolster their chances of winning the match by rotating the strike regularly throughout their chase i.e. routinely taking turns to face the bowler by performing certain cricketing maneuvers such as taking a run. This tactic has notable benefits - it unsettles the rhythm of the bowler thereby reducing the possibility of receiving a wicket-taking delivery and increasing the chances of an easy scoring opportunity. Moreover, it also allows a batsman to get off-strike and renew his energy and focus. All in all, the partnership would flourish and momentum would be preserved throughout the daunting chase.
In terms of Lean Manufacturing objectives, this is analogous to Levelling of Production Mix which entails that the production of all the products in a Product Family should be evenly distributed across the available working time i.e. a process should Changeover frequently into producing a different product rather than continuously producing one product for a long time, regardless of it being in small quantities at a time.
The benefits to be derived from Levelling the Production Mix are several-
it promotes Production flexibility - helping to respond quickly to fluctuations in demand/orders
it reduces the need for keeping excess Inventory as the production of all the products in the Product Family would be more frequent
it reduces the Production Lead Time by virtue of having less of excess/stagnant Inventory
it could reduce the rate of Defects triggered from manufacturing large volumes complacently
it improves the focus, performance and wellbeing of operator resources through the development of a stable and predictable production environment
Note: Levelling the Production Volume and Levelling the Production Mix are deployed together to facilitate Lean Manufacturing - the overarching technique is known as Heijunka/Load-Levelling whose objective is to generate predictability in production operations despite demand volatility.
To summarize, using this ODI Cricket Chase analogy, I've conveyed two important aspects of Lean Manufacturing, which are also critical while designing a Future State Value Stream-
Rather than issuing large Production targets to all manufacturing processes, small targets should be issued at Pitch intervals (derived using a production and a consumption parameter) throughout the workday, and to preferably to just one manufacturing process which will become the Pacemaker of the Value Stream - typically this is the most downstream (customer-facing) process (there can be exceptions to this norm, particularly in manufacturing organizations making customized products where it becomes necessary to regulate the Material Flow at a process further upstream i.e. at an earlier stage in the Value Stream. Refer the TWI Industries video case which has this quirk). This approach is known as Levelling the Production Volume. The processes upstream to the Pacemaker should attempt to synchronize their respective rate of production to the rate of consumption of their downstream counterpart i.e. tether themselves to one another via Pull. This would foster teamwork; the process operators would realize how their efforts are directly contributing (TAKT Image) towards the fulfillment of customer objective
Rather than producing one product in a Product Family for a long period of time, regardless of it being in small quantities at a time, before switching over into producing another product, a manufacturer must strive to evenly distribute the production of all the products in a Product Family across the available working time by performing frequent Changeovers - this technique is known as Levelling the Production Mix. There are several benefits to be reaped, primarily that it would enhance flexibility in Production allowing the manufacturer to respond effectively to demand volatility or last-minute changes in customer orders besides reducing the generation of excess Inventory among other types of Waste
4.6 GUIDELINES FOR DESIGNING OF A LEAN MANUFACTURING VALUE STREAM
Slider 1: Current State and Future State Value Stream Map compared (zoomable)
Figure 25 below summarizes the key objectives to incorporate Lean Manufacturing through the Future State Design of a Value Stream (as alluded to using the three analogies just prior).
The upcoming sections correspond to the sequence of the 8 Key Questions for Future State Design.
4.7 INTERPRETING THE CHANGES PROPOSED IN ACME'S FUTURE STATE MAP
4.7.1 SPOTTING THE BOTTLENECK PROCESS(ES)
To identify the bottleneck(s) in Acme's Steel Brackets Value Stream, if any, let me first derive the TAKT Time which is the minimum rate at which Acme must produce Finished Brackets in order to fulfill its customer demand within the available working time. Acme has established that State Street's monthly demand is expected to remain the same in the future - 18400 Finished Brackets (12000 RH + 6400 LH). Since there 20 workdays in a month, the average demand per workday translates to 920 Finished Brackets. Given that 20 minutes in a Shift is apportioned towards a lunch break, the available working time per workday i.e. across 2 Production Shifts is 55200 seconds. Hence, the TAKT Time for the Steel Brackets Product Family is 60 seconds (available working time ÷ average customer demand).
This TAKT derivation essentially means that if Acme manufactures Finished Brackets at a rate not slower than one-per-minute when operational, it will be able to fulfill State Street's monthly demand.
The Cycle Time data of the manufacturing processes captured within the Current State Map (Figure 26 below) indicates that of the five manufacturing processes, only Assembly 1 is producing output at a pace that is slower than TAKT, which makes it the only bottleneck in Acme's Value Stream currently.
4.7.2 INTEGRATING CONTINUOUS FLOW IN THE VALUE STREAM
Integrating manufacturing processes with Continuous Flow, wherever pragmatic, is a top priority in Lean Manufacturing. Below is a conceptual explanation of this Production Strategy-
CONCEPT NOTE: CONTINUOUS FLOW PRODUCTION STRATEGY
Imagine that there are three sequential activities occurring - Drilling, Coating and Inspection - in place of Assembly 1 at Acme. Each of these activities is performed by a single operator.
On the Value Stream Map, these activities are clubbed together in a single Process Box - separate Process Boxes are not allocated to each of them as material does not stagnate anywhere between these activities i.e. the Inventory is in a state of being continually processed. This can be attributed to the fact that all the operators are seated within a compact workstation (refer Figure 28) where the drilled output is instantly transferred to the coating operator who begins processing it immediately and similarly, the coated output is instantly transferred to the inspection operator who begins processing it immediately.
Continuous Flow is also referred to as One Piece Flow as the Inventory is processed and transferred from one activity to the next one-unit-at-a-time i.e. individually (Batch transfer would invariably induce stagnation).
Note: One can make an allowance for occasional and/or tiny Inventory stagnation in a Continuous Flow setup due to lapse in synchronization, particularly if the activities are performed manually (without mechanization).
The inspected material is transferred to Assembly 2, however, it stagnates at this juncture prior to being processed. Hence, Continuous Flow breaks after Inspection and Assembly 2 is depicted as a separate Process Box.
4.7.2 Continued
The sequence of vetting the processes which can be integrated in a Continuous Flow setup begins with the most downstream manufacturing process (i.e. from right to left in a Value Stream Map), the rationale being that this process lies closest to the customer (akin to the mouth of the bottle in the Bottle Neck analogy) and would need to churn output at a pace that fulfills customer demand with the highest degree of reliability. In general, any Lean Manufacturing endeavour that significantly affects the entire manufacturing operations (Continuous Flow being of fundamental importance) is contemplated keeping the customer objectives in mind. So let me begin evaluating Assembly 2...
Wait! shouldn't Shipping department be the first process to be evaluated for Continuous Flow?
Excellent question but the answer is No, the rather blunt rationale being that as Continuous Flow is a Production Strategy and Shipping department does not perform any Production activity. That being said, allow me share a more nuanced explanation because technically, the Shipping department can be integrated with Continuous Flow, provided an acutely rigid condition is met. Shipping is the most downstream customer-facing process of any Value Stream - linking it in a Continuous Flow setup would imply that Inventory must not stagnate between Production and Outbound Transportation i.e. the moment a Finished Good is produced, it is immediately loaded onto a transportation vehicle. However, as you would concur, this virtually never happens as the Shipping department needs to stage the Finished Goods first - Staging entails the use of a temporary storage area to accumulate the Finished Goods for sorting and/or manifesting prior to being sanctioned for dispatch, subject to the transportation vehicle also being available. Hence, it is a given that Inventory will stagnate at this juncture and thus, by definition, Shipping department cannot be integrated with Continuous Flow.
Note: While stagnation of Inventory is inevitable at this juncture, a manufacturer can still regulate the quantity of it - as elaborated in a later subsection.
With this aspect clarified, let me begin evaluating Assembly 2 process (in conjunction with the Assembly 1 process preceding it as Continuous Flow entails linking two or more activities).
The first question to ponder is - is there anything to suggest that linking these activities together in a Continuous Flow workstation would not be viable or pragmatic?
First, let us consider the characteristics of the Steel Brackets Product Family at Acme - neither are there many products to contend with (just two) nor are the manufacturing processes too complex (just a few, standard operations). Therefore, there should not be any hesitation to designate the most downstream manufacturing process as the Pacemaker of Acme's Future State Value Stream, one which shall receive Production instructions to manufacture small batches of output throughout the workday from the Production Control department.
Integrating the Pacemaker process with Continuous Flow, provided it can churn out material at the pace of TAKT, is ideal from a Lean Manufacturing perspective as this Production Strategy facilitates the purest form of Pull i.e. all the activities within the workstation will be perfectly synchronized - the upstream activity will produce, with maximum reliability, only as much output as the consumption of the activity downstream, resulting in zero Inventory stagnating in-between.
Therefore, what we need to assess is whether Assembly 2 and Assembly 1 can work together to produce output at the pace of TAKT which is 60 seconds per Finished Bracket at Acme. Observe from the Current State Map that Assembly 2 is producing Finished Brackets at a rate that is 20 seconds quicker than TAKT - while this process is not a bottleneck, however, it is overproducing by a big margin which is not ideal, especially considering that it is a customer-facing process.
Shouldn't this be easy to resolve - can't we have Assembly 2's operator perform his job slowly?
No, promoting inefficiency is counteractive to Lean Manufacturing. Instead, it will be prudent to optimize the operator utilization (elaborated in a Rapid Process Improvements subsection).
As for Assembly 1 process, recall that it is a bottleneck currently as it is producing output at a rate that is slower than TAKT and thus, Acme would need to find a way to improve its efficiency, which shouldn't be all that difficult to do as the deviation from what is desired is just 2 seconds/~3%. The operator can perhaps be trained to perform the assembling activity quicker.
As such, there aren't any major impediments hindering Assembly 2 and Assembly 1 processes from being considered to be integrated with Continuous Flow. It helps to know that both these processes have an Uptime of 100% as well as take no time to perform a Changeover which means that they are perfectly reliable and joining them wouldn't be risky. If the implementation is successful, then the excess Inventory that currently stagnates between these processes (1200 LH and 640 RH Brackets) would be completely eliminated, freeing up 2 days from the Total Production Lead Time.
The next question to consider is - Is there any scope to extend the Continuous Flow to more activities?
It is evident from the Current State Map that significant quantity of Inventory is stagnating before all the manufacturing processes upstream as well - between Spot Weld 2 and Assembly 1, between Spot Weld 2 and Spot Weld 1 and between Stamping and Spot Weld 1 processes. Having all of them integrated with Continuous Flow, provided it is viable, would ensure that the entire manufacturing operations is tethered by pure-Pull with zero overproduction and zero stagnating Inventory anywhere - the Total Production Lead Time would shrink to just 5 days and 188 seconds ( an ~80% reduction).
Wouldn't this Value Stream be too idealistic, one that struggles to withstand the rigours of reality?
Excellent fodder for a riveting debate, however, we may be getting too ahead of ourselves. Instead, let' continue evaluating the remaining processes, utilizing the objective information captured in the Current State Map along with our subjective understanding about the nature of the activities involved.
For starters, it is known that none of the remaining processes are bottlenecks. In fact, all of them produce output much quicker than TAKT, which is addressable as I had indicated earlier while evaluating Assembly 2. The Assembling activities are preceded by Spot Welding activities - while the nature of both these set of activities are different, the complexity involved isn't - both are rather standard, straightforward operations, which is conducive for Continuous Flow. That being said, Spot Weld 2 is not perfectly reliable (80% Uptime) unlike Spot Weld 1, and this would certainly hamper the synchronization at the workstation laying to waste the very purpose of its existence. It is a no-brainer that only if this issue can be remedied should Spot Weld 2, and by extension Spot Weld 1 which lies further upstream, should be integrated within the workstation. Thankfully, Acme is confident of being able to do so - elaborated in a Rapid Process Improvements subsection.
Also, note that both the Spot Welding activities take some time to perform a Changeover (10 minutes). This is not a major issue per se as Changeover Time, which is often inevitable and necessary, can be counterbalanced by producing output proportionately quicker than TAKT, which these processes already seem to be doing. Nonetheless, reducing the time taken to perform a Changeover would be worthwhile in any case (and which Acme shall pursue) as unnecessary Waiting Time and Motion, which often tend to be a significant constituent of Changeover Time, are types of Waste and Waste, as you already know, should be eliminated from the Value Stream as much as possible. All in all, having Spot Weld 1 and 2 within the same Continuous Flow workstation alongside Assembly 1 and 2 should be alright.
Prior to Spot Welding 1 lies the last manufacturing process pending to be evaluated - Stamping. The 200 T Stamping Press equipment that this process utilizes should capture your attention - the nature (mechanical) and complexity (high precision) involved in this activity is starkly different from Spot Welding and Assembling activities, and this would be a deterrent to Continuous Flow. Also, it churns out Stamped Brackets at rapid speed - its Cycle Time is just 1 second. To slow this process down by upto 59 seconds to make it compliant with its downstream counterpart - Spot Weld 1, and by extension, TAKT does not seem very practical. Stamping Press is also not perfectly reliable ( 85% Uptime) and this would be damaging to a Continuous Flow workstation. However, Acme believes it can remedy this issue through a rapid Process Improvement intervention.
However, there is another factor which dispels all doubts and conclusively rules out Stamping from being considered - the Stamping Press equipment operates on multiple Product Families at Acme i.e. it does not exclusively manufacture Steel Brackets alone (this aspect is already captured in the Current State Map through the cross-markings on Stamping's Process box). If one were to integrate Stamping in the Continuous Flow workstation which would be producing output at a rate similar to TAKT, the Stamping Press would exhaust virtually all its available working time on producing Stamped Brackets alone, leaving the other Product Families bereft of its services. Acme would also be compelled to procure/lease another one of this large and expensive equipment, which would be a highly imprudent thing to do. As a result, it has decided that the Continuous Flow workstation will only perform both the Spot Welding and both the Assembling activities. From a Value Stream Mapping perspective, this new and discrete manufacturing process shall be named Weld+Assembly.
4.7.3 IF NOT CONTINUOUS FLOW, THEN WHAT?
While Continuous Flow is in many ways ideal for Lean Manufacturing, there are instances where it is either impractical or plain unviable to deploy this Production Strategy - as was the case for Acme's Stamping process. Manufacturing processes that are affected by demand/supply volatility, transportation delays and/or equipment downtime would actually benefit from having excess Inventory before them to serve as a shield against such disruptions. Regardless, it is essential that a manufacturer attempts to regulate the quantity of the excess as it still is a Waste in the system. To do so, it can make use of alternatives strategies such as Supermarket Pull System and FIFO (First-In, First-Out) Lane. Let me share a conceptual background for these first prior to elaborating where and how Acme can benefit from deploying either/both of them in its Future State Value Stream.
CONCEPT NOTE: THE SUPERMARKET PULL SYSTEM
A Supermarket Pull System entails the use of a Supermarket for regulating the quantity of excess Inventory and Kanban to synchronize Production with Consumption through a robust information transmission mechanism that works reliably in a high-transaction-frequency environment.
SUPERMARKET
At a traditional Supermarket, a customer walks through the aisles, collects the desired products from the shelves and pays for them at the billing desk. The staff monitor the shelves at regular intervals and replenish the ones that need attention. A Supermarket in a Lean Manufacturing setup is not too different - think of it as a regulated Inventory zone where a consumer process obtains its material inputs based on its immediate requirement. Correspondingly, or as soon as the quantity of Inventory falls below a set threshold, instructions are issued to the manufacturing process upstream to produce a new batch of output and replenish the Supermarket Inventory to its default level. Thus, this facility ensures that Production is tethered to Consumption via Pull, which serves to regulate the quantity of excess Inventory in the Value Stream.
KANBAN
Just regulating the Material Flow isn't enough. Kanban is a nifty signalling system which helps secure the Information Flow. Recall that in a Lean Manufacturing environment, Production instructions would be issued to the Pacemaker process in small quantities throughout the workday and the manufacturing processes upstream would need to align their Production to the Consumption of their downstream counterpart. This implies that the frequency of information transmission would be very high across the Value Stream and therefore, it would be vital that the method is is easy-to-transmit, easy-to-interpret and easy-to-audit. All of these can be reliably facilitated through Kanban.
The graphic below depicts a Supermarket Pull System deployed at a Computer Hardware manufacturer which I will use to explain how Kanban works. It would help if you would recall the Library Card System analogy here as the the Kanban components (Figure 31) have a similar function within a Supermarket Pull System as the Library Cards (Figure 19) have at a Library.
Note: Just as the Library Cards System, the use of Kanban is increasingly being done electronically.
The operator at Process B wants to begin manufacturing his next batch of output and proceeds to fetch the input material, a Tray containing 20 A-chips, from the Stock Bin adjacent to his workstation which he empties on his desk. A patrolling material handler spots the empty Tray and pulls the attached Withdrawal Kanban Card from it and subsequently visits the Supermarket to deposit it which signals to the Supermarket staff that Process B has consumed a Tray of 20 A-chips and that the Inventory in the latter's Stock Bin needs to be replenished. The staff fetches a new Tray containing 20 A-chips from the Supermarket shelf and gives it to the material handler who visits Process B's workstation and deposits it within the Stock Bin. This closes the Inventory Withdrawal loop.
At the Supermarket, the staff proceeds to print a Production Kanban Card and hands it over to another patrolling material handler who visits Process A and submits it to its operator. The operator interprets the receipt of this card as a signal to initiate manufacturing a new batch of 20 chips. Once the output is ready, the material handler will transfer it to the Supermarket, thereby replenishing its Inventory to its original level. This is the closure of the Inventory Production loop.
Thus, the utility of the smart information signalling system that Kanban is, is that it tethers Production to Consumption (fundamental for Lean Manufacturing) in a quick and reliable manner ensuring that the instructions are easy to interpret and conform to for all the stakeholders involved.
The other components of Kanban are utilized in specific situations only-
While Lean Manufacturing detests Production to occur in large batches, exceptions may need to be made wherever pragmatic (as Acme would realize as well) - for example if a particular manufacturing process situated at another factory. While the Inventory Withdrawal loop of Kanban may continue to function as intended i.e. consumption of inputs in small quantities represented by a Withdrawal Kanban card of that denomination; however, the Inventory Production loop can be altered to accommodate an unusual instruction to produce a large batch. In such a situation, the Supermarket cannot issue a Production Kanban Card (which must correspond to the denomination of the Withdrawal Kanban Card). Also, issuing multiple Production Kanban Cards together totalling up to the large production quantity wouldn’t exactly be fail-safe information signalling. Instead, using a Signal Kanban that represents this predetermined large production quantity would be more effective. The object utilized as Signal Kanban is typically a metal triangle and it will be issued to the producing process when the quantity of Supermarket Inventory reduces by the same amount as the size of the large batch to be produced.
As for the Kanban Post, it has a utility that is similar to that of the Book's Card Holder in the Library analogy. The Kanban Post is a receptacle situated near the consuming process where the Withdrawal Kanban Cards accumulate prior to being transferred to the Supermarket. This is an unconventional approach as a Withdrawal Kanban Card is typically transferred individually. Functionally, the Kanban Post is to the Inventory Withdrawal loop what Signal Kanban is to the Inventory Production loop - the latter is utilized when there is a unique production need, the former is utilized when there is a unique withdrawal need - for example, when it is pragmatic to withdraw a large quantity from the Supermarket. Acme has utilized Kanban Post in its Value Stream as well, for another valid reason.
The Material Pull component is just an icon to be utilized during Value Stream Mapping if the Inventory is being transferred from the Supermarket physically i.e. without the aid of material handling equipment such as Forklifts or Trolleys.
To summarize, the Supermarket Pull System is a potent technique which facilitates Lean Manufacturing by combining the utility of Supermarkets (regulation of Inventory) with that of Kanban (robust Information signalling which synchronizes Production to Consumption). This system is typically deployed if the number of products in a Product Family are few and/or if their Production Lead Time is short. The rationale is simple - maintaining Inventory for multiple products in a high transaction frequency environment is tough. Also, if the Inventory cannot be manufactured in quick-time, it would jeopardize the very foundation of a Supermarket which relies on quick replenishment against withdrawals.
CONCEPT NOTE: FIFO (FIRST-IN, FIRST-OUT) LANE
FIFO (First-In, First-Out) Lane is another alternative Lean Manufacturing strategy which is utilized when deploying a Supermarket Pull System is not practical but the need to regulate the production of excess Inventory is still essential. Let me explain the two terms which make up this technique-
First-In, First-Out is an Asset Management Strategy which stipulates that Inventory must be consumed based on the order that it was produced i.e. from oldest to newest (LIFO or Last-in, First-Out strategy stipulates the opposite).
Lane: Refer the demonstration in Figure 36 below where Process B (consumer) is located on the ground floor and Process A (producer) lies directly overhead on the first floor of the factory. A hollow Tube (representing the Lane) connects both these floors and can store up to six cartons of Process A's output/Process B's input. The operator of Process B will extract cartons from the Tube as per his immediate requirement while the operator of Process A would receive a visual cue as to how much to produce so as to replenish the Tube to its full capacity. Thus, the Tube serves as a repository of Inventory, akin to the Supermarket, while the visual cue ensures Production is linked to Consumption, akin to Kanban. Also, the carton that was produced first would drop to the bottom of the Tube while the next one would stack above it. As a result, it will always be withdrawn by Process B in an oldest-to-newest sequence, thereby ensuring FIFO.
While I have used gravity to illustrate the concept of FIFO, in a standard manufacturing setup the Lane is set up horizontally between the producer and consumer process with FIFO being triggered just like at a travelator in an airport.
A FIFO Lane is easier to maintain than a Supermarket and is typically deployed if there are many products in a Product Family, if their consumption is infrequent, if their Production Lead Time is long and/or if they are perishable or of high value.
4.7.3 Continued
At Acme, the number of products in the Steel Brackets Product Family are just two - LH and RH, and the Raw Material that goes into its production is just one - Steel Coils. The customer demand for Finished Brackets is large which induces Acme to run its manufacturing operations across 2 Shifts every day. Also, Steel Brackets are neither perishable nor expensive products. While the Total Production Lead Time currently is long at ~24 days, the Total Processing Time component of it is miniscule - ~ 3 minutes. All these characteristics suggests to Acme that deploying a Supermarket Pull System would be a better alternative to using a FIFO Lane at those junctures in the Value Stream where Continuous Flow cannot be integrated, subject to viability.
Figure 37 depicts the current Material Flow within Acme - you already know that a Continuous Flow workstation will integrate the four processes from Spot Weld 1 to Assembly 2. Hence, Acme now needs to determine whether to install a Supermarket Pull System at either/all of the three remaining junctures (marked as red circles) i.e. whether to regulate the production of Steel Coils, Stamped Brackets and Finished Brackets in the Lean Value Stream of the future.
While it would be beneficial to regulate the sizeable quantity of stagnating Raw Material Inventory (5-days-worth of Steel Coils) before Stamping process, the decision to deploy a Raw Material Supermarket would hinge on whether the supplier, Michigan Steel can reduce its Delivery Lead Time (deliveries occurs twice-a-week currently) so that it is able to replenish the Steel Coil Inventory at the Supermarket on a daily-basis based on the consumption of Stamping
Having a Stamped Brackets Supermarket would regulate the sizeable quantity of Inventory stagnating before Spot Weld 1 (4600 LH and 2400 RH Stamped Brackets). Given that the Stamping Press churns out output rapidly, Stamping process will be easily able to replenish the Stamped Brackets Supermarket based on the consumption pattern of Spot Weld 1, however, this is subject to how frequently the Steel Coils can be delivered to it by the supplier in the first place
Having a Finished Goods Supermarket would regulate the sizeable quantity of Inventory stagnating before the Shipping department (2700 LH and 1440 RH Finished Brackets), however, its viability is once again dependent on whether the Production Lead Time of the processes upstream is short enough to replenish this Supermarket based on the rate of consumption of the Shipping department. Alternatively, Acme also has the option to have the Pacemaker process Weld+Assembly produce directly to Shipping instead of producing to a Supermarket i.e. it can deposit the Finished Brackets at the Staging zone as soon as they are produced. Doing so would eliminate the need to have this Supermarket (and the excess Inventory associated with it), however it would also expose the Shipping department to disruptions such as production delays and demand volatility which would not be ideal, especially considering that Acme will be adapting to a Lean Manufacturing Value Stream for the first time. The customer-facing manufacturing process (the Pacemaker - Weld+Assembly) in particular will be pursuing Process Improvement initiatives in order to become a reliable Continuous Flow workstation commanding Acme's Material Flow, as indicated in the previous section. Hence, all things considered, it will be prudent for Acme not to consider producing directly to Shipping
SUPERMARKET 1: STEEL COILS (RAW MATERIAL) SUPERMARKET
The Steel Coils Supermarket would serve to link Michigan Steel's production to Stamping's consumption via Pull.
Acme has determined that it can have the Steel Coils delivered to Stamping once every day in the future (twice-a-week currently). Michigan Steel has several customers in this region and Acme believes it can convince it to adopt a Milk Run delivery model which entails dispatching a truck carrying small consignments for multiple customers daily rather sending a truckload to a single customer less frequently as it does today.
What is in it for Michigan Steel? Would it stand to benefit by making this adjustment?
Certainly. Its customers would be receiving more frequent deliveries at no added cost which would be a major service improvement. Besides, its manufacturing unit would continue operating as usual.
Therefore, Acme has decided to install a Raw Material Supermarket and keep 1.5 days of Stamping process' Steel Coil consumption as Inventory within.
The 1.5 days constitutes Cycle Inventory to cater to Stamping's daily consumption until a new consignment arrives next day and additional half-a-day of Buffer Inventory to guard against unexpected and occasional issues such as sudden surges in customer demand, transportation delays or material defects. By introducing a Raw Material Supermarket, Acme will be able to reduce its Production Lead Time by 3.5 days at this juncture (5 days of Inventory stagnating currently − 1.5 days of Supermarket Inventory), which is a 70% reduction.
Note: The Inventory kept in a Supermarket is not to be mistaken with Safety Stock - the latter is utilized temporarily in emergency situations such as if a manufacturing equipment breaks down or if there is a disruption in the supply chain. Safety Stock is stored and accounted for separately as well.
In terms of Information Flow, as there is an external entity involved (the supplier), Kanban will not be utilized in the traditional way. Rather, the Withdrawal Kanban Cards corresponding to the Inventory utilized by the Stamping process (each card would denote the consumption of a single Steel Coil) would accumulate in a Kanban Post. At the end of each workday, Production Control department will arrange to collect the bunch of accumulated Cards, calculate the total consumption for the workday and issue the corresponding Purchase Order (not Signal Kanban) to Michigan Steel, who in turn will replenish the Supermarket Inventory to its default by dispatching a consignment the next day.
SUPERMARKET 2: STAMPED BRACKETS SUPERMARKET
The Stamped Brackets Supermarket would serve to link Stamping's production to Weld+Assembly's consumption via Pull.
Given the positive repercussions from installing Supermarket 1 (daily delivery from supplier) as well as Stamping's own rapid Cycle Time of 1 second, it is a no-brainer for Acme to install a Stamped Brackets Supermarket to regulate the excess Inventory stagnating at this juncture.
Just like the Raw Material Supermarket, 1.5 days of Inventory will be kept at this Supermarket - Cycle Inventory to cater to Weld+Assembly's daily material consumption and additional half-a-day of Buffer Inventory to guard against unexpected and occasional issues. By introducing the Stamped Brackets Supermarket, Acme will be able to reduce its Production Lead Time by 6.1 days at this juncture (7.6 days of Inventory stagnating currently − 1.5 days of Supermarket Inventory), which is an ~80% reduction.
In terms of Information Flow, Kanban will be utilized in an unconventional way at this Supermarket as well - while the Withdrawal Kanban will be transferred as default (individually) to the Supermarket, where each Card would reflect the consumption of one Bin of Stamped Brackets (Bin is a plastic container containing 60 Stamped Brackets of the same type), this equivalence shall not be maintained for the Inventory Production loop i.e. a corresponding Production Kanban Card to manufacture 60 Stamped Brackets will not be issued to the Stamping process. Instead, this Supermarket will issue a Signal Kanban which will represent an instruction to produce a predetermined large batch quantity - 300 LH or 160 RH Stamped Brackets (interchangeably).
What is the logic behind this unusual instruction to manufacture a large batch of output?
Before addressing this question, let me emphasize that the denomination of Withdrawal Kanban - a Bin of 60 Stamped Brackets - represents one hour of processing work for the Pacemaker Weld+Assembly process if it were to produce to TAKT as recommended by Lean, which in Acme's case is 60 seconds i.e. one-Finished Bracket-produced-every-minute. Thus, upon receiving the Withdrawal Kanban, the Supermarket would supply another Bin of Stamped Brackets to Weld+Assembly i.e. another hour-worth of input material. Also, this 60 Stamped Brackets denomination is a derivative of the Pack Size (3x) - recall that State Street orders Finished Brackets in Trays, each of which represents 20 exclusively LH or RH Finished Brackets. That the denomination of Withdrawal Kanban is linked to an important production parameter like TAKT and an important customer parameter like Pack Size will be handy while performing Production Volume Levelling as indicated earlier (also explained in detail in an upcoming section).
Now here's the important part - as for the Inventory Production loop, transferring a corresponding Production Kanban Card to Stamping process, which would be an instruction to manufacture 60 Stamped Brackets of a particular kind, would not be practical as given Stamping's Cycle Time of 1 second, it would finish manufacturing this small target in a minute. This is too tiny a denomination of work, even by Lean Manufacturing standards. You'll also have to consider that the operator at the Stamping process would also need to perform a Changeover frequently i.e. set up the Stamping Press equipment into manufacturing another product from the Product Family - this is due to the application of Production Mix Levelling as indicated earlier (also explained in detail in an upcoming section). The Changeover Time for Stamping is very high currently (1 hour) and regardless of any possible improvement that could be made to it, performing a Changeover after every two minutes of producing LH Stamped Brackets or one minute of producing RH Stamped Brackets (as the ratio of LH:RH demand is 2:1) would not be pragmatic at all. Stamping Press also has to serve other Product Families at Acme as well - it is not utilized towards manufacturing of Steel Brackets alone.
Hence, issuing a Signal Kanban, which serves as an instruction to manufacture a large batch of 300 LH Brackets or 160 RH Brackets after performing a Changeover, would be a much more practical Production instruction. The logic behind the denomination is that as State Street's rate of demand for Finished Brackets is 600 LH and 320 RH per day, this would make the Signal Kanban an instruction to produce an entire Shift-worth of LH or RH Stamped Brackets (recall that Acme operates 2 Shifts in a workday), thereby making it a practical instruction to conform to for the process operator.
The Supermarket will issue a Signal Kanban whenever the Cycle Inventory of Stamped Brackets falls by 5 LH Stamped Bracket Bins or 3 RH Stamped Bracket Bins i.e. upon accumulating 5 LH Withdrawal Kanban Cards or 3 RH Withdrawal Kanban Cards from the consuming Weld+Assembly process (which would be transferring these individually).
SUPERMARKET 3: FINISHED BRACKETS SUPERMARKET
The Finished Brackets Supermarket would serve to link Weld+Assembly's production to Shipping's consumption via Pull.
Given the positive repercussions on the Production Lead Time from the installation of Supermarket 1 and 2 (9.6 days eliminated) and the benefits to be derived from installing the Continuous Flow workstation, Acme is in no doubt about the utility of installing a Supermarket at this juncture.
Besides regulating the stagnation of Finished Brackets Inventory, this Supermarket will also serve another purpose - it will shield the Pacemaker Weld+Assembly process by bearing the brunt of demand volatility itself - Acme has decided that it will keep 2 days of Finished Brackets Inventory at this Supermarket - Cycle Inventory to cater to the Shipping department's daily consumption and additional one-day of Buffer Inventory this time to hedge against sudden surges in demand and last-minute changes in customer orders besides other unexpected and occasional issues. By installing a Finished Goods Supermarket, Acme will be able to reduce its Production Lead Time by 2.5 days (4.5 days of Inventory stagnating currently − 2 days of Supermarket Inventory), which is a 56% reduction.
In terms of Information Flow, Kanban will be utilized as default - both Withdrawal and Production Kanban will be transferred individually and its denomination will be 20 Finished Brackets (practical as it corresponds to the customer's Pack Size of a Tray).
4.7.4 LOAD LEVELLING WITH PACED WITHDRAWAL AT THE PACEMAKER
The groundwork has been laid to usher in Lean Manufacturing and the next step would be to optimize the Material Flow through this redesigned Value Stream.
This essentially entails just determining the Production instructions that need to be issued to the Pacemaker Weld+Assembly process (more precisely to the Spot Welder 1 within), as the optimized flow will automatically extend upstream to all the other manufacturing processes due to Pull facilitated by the three Supermarkets.
CONCEPT NOTE: LOAD LEVELLING WITH PACED WITHDRAWAL
LOAD LEVELLING
You may recall the Cricket analogy now, where I had introduced Load Levelling, known as Heijunka in Japanese, and shared the rationale for performing this technique.
The Load comprises of two constituents - Production Volume and Production Mix - and Levelling the Load entails deriving how much to produce and how to sequence the production respectively, in order to incorporate Lean Manufacturing in the Value Stream in the best possible manner.
The purpose of Levelling the Production Volume is to miniaturize the Production target so that it can be issued at regular intervals which will generate predictability in the Material Flow (Lean detests unevenness - refer the Japanese concept of Mura). For example, if the Forecast for Finished Goods next week is 1000 units, the idea is not to issue a large Production Schedule stating produce 1000 units over the next week or even produce 700 units by Thursday and a further 300 by Sunday. Such bulky and/or irregular targets will only incentivize the process operator to produce the target quantity in large batches, and by doing so, will lose his sense of TAKT Image and untether himself from the objective of producing to the customer’s rate of demand, in the process becoming susceptible to Overproduction, increase in Defects and other types of Waste.
To what extent should the Production Instructions be miniaturized?
The miniaturized Production Instruction or Pitch should be a practical quantity that is a derivative of an important production parameter - typically the TAKT Time, and an important customer parameter such as Pack Size or Minimum Order Quantity or Average Order Size. For example, if Mapro is capable to produce to a TAKT of 60 seconds (a-bottle-a-minute) and the customer transacts in cases where each Case contains 30 Syrup bottles, then the miniaturized Production instruction or Pitch can be practically set to 30 minutes or 1 hour (TAKT x Pack Size, or TAKT x Pack Size x 2) i.e. an instruction to manufacture 30 Syrup bottles or 60 Syrup bottles in a batch.
Note: The Pitch is transmitted using Kanban and not in a written format.
Figure 43 below demonstrates visually, using another example, how Levelling the Production Volume helps to safeguard manufacturing operations from fluctuating demand-
A stable flow of Production instructions reduces stress for the process operators. This is in contrast to Batch Production where encountering demand volatility would induce the process operators to work overtime on some days while being idle on others.
What if the produced quantity of Finished Goods in a day along with excess Inventory from the previous days is not sufficient to meet the customer demand for that particular day?
That is certainly a possibility and an unfavorable one at that, however, this is precisely why a Finished Goods Supermarket keeps Buffer Inventory - to safeguard against this risk. If demand volatility is anticipated to be significant, then the quantity of Buffer Inventory kept should be higher - Acme allocated an extra 0.5 days of Buffer Inventory for its Finished Brackets Supermarket compared to its other two Supermarkets for this very reason - as this Supermarket faces the customer in the Value Stream and has to be able to bear the brunt of sudden demand surges or last-minute changes in customer orders so as to shield the all-important Pacemaker process effectively.
The purpose of Levelling the Production Mix is to ensure that all the products in a Product Family are manufactured in an evenly distributed manner across the available working time.
Consider there are 3 products in a Product Family: A, B and C. The daily customer demand for these are 200 units, 400 units and 800 units respectively (1400 in total, the ratio being 1A: 2B : 4C). The Cycle Time is 30 seconds per unit for all the products and this conforms to TAKT. Production occurs for 16 hours in a workday across 2 Shifts. The Changeover Time, i.e. the time taken to setup equipment and resources to transition into manufacturing another product from the same Product Family, is 20 minutes. The customer transacts in cartons (Pack Size) where each Carton contains 20 units. The Average Order size for the customer historically has been 50 products.
An organization can choose to adopt this production sequence to fulfill the customer demand-
200 A → Changeover → 400 B → Changeover → 800 C
This is a classic example of a Batch Production sequence which has traditionally been popular for the very reason you may have seen it favourably - large Volumes and few Changeovers to perform. However, producing large quantities of the same product before changing over leaves little room for flexibility in manufacturing operations and the organization will be unable to respond effectively to either demand volatility or last-minute changes in customer orders. Moreover, producing in large batches is likely to trigger Overproduction and excess Inventory besides other forms of Waste.
However, if the manufacturer were to apply Heijunka/Load Levelling maintaining the current parameters, then the production sequence can be arranged in such a manner-
50 A → Changeover → 50 B, 50 B → Changeover → 50 C, 50 C, 50 C, 50C → Changeover → repeat 3 more times
This sequence was derived by computing how much time is left after deducting the total processing time from the total available working time - the remaining time was allocated towards performing Changeovers and subsequently, production was evened-out around it. As the Cycle Time is 30 seconds, the time taken to produce 1400 units would be 42000 seconds or 11 hours and 40 minutes. That would leave 4 hours and 20 minutes of time, sufficient to accommodate a maximum of 13 Changeovers of 20 minutes each. The manufacturer decided to incorporate 12 Changeovers and distribute the production of A, B and C evenly around it i.e. Levelling the Production Mix. The Pitch is derived to be 25 minutes through Production Volume Levelling (TAKT x Average Order Size) i.e. instructions will be released to manufacture 50 products of a particular type in a single batch regularly throughout the workday. As is evident, Load Levelling/Heijunka utilizes Production Volume Levelling and Production Mix Levelling concurrently.
While this sequence appears complicated due to small Volumes and several Changeovers, however, it would help the manufacturing operations become more flexible, a trait necessary to deal with disruptions such as fluctuations in customer demand. Not only that, Production Lead Time will be shortened and excess Inventory besides other types of Waste will be regulated as well. All these aspects are beneficial from the integrating of Lean Manufacturing in the Value Stream perspective.
Figure 43 below demonstrates visually, using another example, how Levelling the Production Mix helps to instill flexibility in manufacturing operations-
Through Process Improvements, if the time per Changeover decreases (and/or if the Cycle Time reduces), then Load Levelling would be able to accommodate more Changeovers which would lower the Pitch as well as even-out the production sequence further. This highlights an important aspect of Lean philosophy - Continuous (Incremental) Improvement - popularly known by its Japanese equivalent - Kaizen. That being said, as the implementation phase of the Future State Lean Value Stream Design is time-critical, the incremental improvements would need to be accomplished at a much higher frequency than would be acceptable in another, more normal, situation. This would necessitate having rapid Process Improvement interventions instead, known as Kaizen Burst in Lean Manufacturing parlance - explained in detail in an upcoming section.
But how exactly can Changeover Time be reduced, though?
Changeover Time has its constituents and remedial measures exist for each of them-
Time taken to setup the equipment: can be reduced by retooling
Time taken to bring input material to the workstation: can be reduced by keeping Inventory in close proximity to the process operator, preferably within the workstation itself (akin to the Bin utilized in Acme's Stamping Supermarket)
Time taken to perform quality checks: can be reduced by mistake-proofing (Poka-Yoke), through preventive maintenance of equipment, by using faster inspection methods etc.
While certain improvements would necessitate making Engineering or Design changes - particularly for the first constituent listed above, it would help you to know that a significant reduction in Changeover time can be achieved just by adhering to Lean Manufacturing guidelines. For example, in a Continuous Flow arrangement, the process operators would be seated in a compact workstation and all the material required for production will be kept in close proximity - this would help mitigate the second constituent listed above. Besides, the manufacturing processes upstream will supply output on-demand to their respective downstream counterpart, by virtue of Supermarket Pull Systems and/or FIFO Lanes being integrated in the Value Stream. This would result in less Waiting Time for the process operators. Also, by moving away from Batch Production, a manufacturer would cease to overproduce, which is a type of Waste that spurs the generation of another type of Waste - Defects - this would help address the third constituent listed above.
Continuing with the written example, let's say that the manufacturer has reduced the time taken to perform a Changeover from 20 minutes to 8 minutes through a Kaizen Burst. This would allow him to accommodate 30 Changeovers and set an improved Pitch of 10 minutes, a practical quantity which corresponds to the Pack Size this time. The modified Load Levelling production sequence may appear as follows-
20 A → Changeover → 20 B, 20 B → Changeover → 20 C, 20 C, 20 C, 20 C → Changeover → repeat 9 more times
Note: Lean guidelines suggest that a key objective for a manufacturer is to continue increasing the Production Mix Levelling i.e. lower the EPE (every part every...), till an EPE-Pitch is achieved. For example, if a manufacturer changes over into making all the products in the Product Family once every day i.e. EPE-1 day, then he must strive to increase the Production Mix Levelling incrementally to EPE-1 Shift and subsequently to EPE-1 hour, and so on. In the production sequence listed above, the manufacturer is able to lower the Pitch i.e. increase the Production Volume Levelling to 10 minutes, which is ideal for him as it corresponds to the customer's Pack Size. The resulting pattern of production is so evenly spread out that any product in the Product Family can be manufactured by performing a Changeover at a 10-minute interval - thus, the Production Mix Levelling corresponds to an EPE-Pitch (the Levelling can be increased even further to EPE-TAKT, but this is too idealistic for all practical purposes).
PACED WITHDRAWAL
Now that you are familiar with how to balance the quantity and distribution of Production, let me explain how the production sequence derived through Load Levelling is integrated in the Value Stream - this technique is known as Paced Withdrawal.
Figure 45 below depicts a Load-Levelling Box, the device used to facilitate Paced Withdrawal. It is a repository of sequenced Withdrawal Kanban Cards for Finished Goods (this is assuming that the Pacemaker process lies at the downstream-end of the manufacturing operations), arranged by the Production department before the workday begins, based on the production sequence derived through Load Levelling.
The Y-axis of the box depicts the name of the products in the Product Family - A, B and C. The X-axis denotes the time of the workday at Pitch intervals of 10 minutes - the arrangement indicates that an EPE-Pitch is practised. The Changeover Time is zero as there are no gaps between the Pitch intervals i.e. the Pacemaker does not need any time to set up into manufacturing a different product from the Product Family (recall that the same is applicable for the two Assembly activities at Acme).
To demonstrate the technique in action, a material handler from the Shipping department will pick the Withdrawal Kanban Card for Product A from the Load-Levelling Box at 8 am and submit it to the Finished Goods Supermarket. This will result in the Supermarket handing over a 10 minutes production-equivalent quantity of Finished Goods to the material handler who will proceed to stage these products at the customer dispatch warehouse. Meanwhile, the Supermarket will seek to replenish the quantity withdrawn by issuing a corresponding Production Kanban Card to the Pacemaker process upstream, which in turn will produce the instructed output and transfer the same, thereby replenishing the Supermarket Inventory it to its original level. This cycle will be repeated at 8.10 am with the Withdrawal Kanban Card of Product B and at 8.20 am with the Withdrawal Kanban Card for Product C, and so on, throughout the workday.
Thus, the Shipping department paces its Finished Goods withdrawals to synchronize with the rate of customer demand and the Pacemaker process produces corresponding to the consumption rate of the Shipping department. This Production-matching-to-Consumption loop will propagate upstream in the Value Stream wherever a Supermarket Pull System and/or a FIFO Lane is deployed.
4.7.4 Continued
Acme has decided to pursue Production Volume Levelling by setting a Pitch of 20 minutes (TAKT x Pack Size). This translates to a Production instruction of 20 Finished Brackets - which is a miniature quantity as well as is practical to align to for the process operators.
As for the Production Mix Levelling, since the ratio of State Street's demand for Finished Brackets is roughly 2:1, Acme has decided to evenly-distribute its production in this manner-
LH, LH → Changeover → RH → Changeover → repeat
Overall, through the application of Load Levelling, the following production sequence will be issued to the Pacemaker Weld+Assembly process in the future-
20 LH, 20 LH → Changeover → 20 RH → Changeover → repeat 14 more times in a workday to fulfill the daily customer demand of 600 LH and 320 RH Finished Brackets.
A total of 30 Changeovers would have to be performed by the Pacemaker daily. As is evident from the Current State Map, the Spot Welders operating from this Continuous Flow workstation would need 10 minutes each to perform a Changeover while the Assemblers do not need any Changeover Time. At this rate, a total of 5 hours will be consumed by a Spot Welder on performing Changeovers alone from the available working time of 15 hours and 20 minutes (16 hours across 2 Shifts − 2 Lunch Breaks totalling 40 minutes), leaving 10 hours and 20 minutes for actual Production.
Can the Pacemaker fulfill the customer demand in this timeframe?
If Weld+Assembly churns out Finished Brackets at the rate of TAKT (60 seconds) as is recommended in the Lean guidelines, then naturally all the available working time will be exhausted producing the 920 Finished Brackets demanded by State Street, leaving no time to perform Changeovers - which is an impossible situation. Hence, Acme would need to find ways to-
either completely eliminate the 10-minute Welding Changeover Time
or churn out Finished Brackets at a Cycle Time of ~40 seconds (Available Working Time of 55200 seconds − Total Changeover Time of 18000 seconds ÷ 920 Finished Brackets)
or a combination of these two options i.e. reducing both, Cycle Time as well as Welding Changeover Time, to a certain extent
The first option would be ideal, however it may not be plausible. Adopting the second option would mean a significant deviation from TAKT, which would contribute to a loss of TAKT Image and trigger the generation of Waste and hence, it would not be preferable. Therefore, at first glance, the most realistic solution would be a hybrid of these two options i.e. Option 3. What Acme does eventually is elaborated in the Kanban Burst section up next.
As for the modality of Paced Withdrawal, a Withdrawal Kanban Card will be drawn by the Shipping department's material handler at Pitch intervals of 20 Minutes from the Load-Levelling Box (refer Figure 46 below) - this is assuming that Changeover Time is completely eliminated. Upon submitting it to the Finished Goods Supermarket, 20 Finished Brackets (either LH or RH) will be given to him which he will deposit at the Staging zone within Shipping. Correspondingly, the Supermarket shall issue a Production Kanban to Spot Welder 1 at the Pacemaker Weld+Assembly process, who will produce and transfer 20 Finished Brackets to it, thereby replenishing its Inventory to the default state.
4.7.5 RAPID PROCESS IMPROVEMENT INTERVENTIONS
The Future State Design of the Value Stream is nearly ready - just a few process improvements would be needed, to be accomplished in quick-time, before we can be assured of having integrated Lean into the manufacturing operations of Steel Brackets at Acme. Let me share the Concept Note first-
CONCEPT NOTE: KAIZEN BURST
There are two ways to ascertain which processes need to be improved, and the nature of improvement that needs to be made-
by spotting inefficiencies in the Current State Value Stream
by fine-tuning the Future State Value Stream Design
As indicated earlier, Lean is based on the philosophy of Kaizen or Continuous Improvement, which entails striving to be incrementally better, perpetually. Rather than making radical changes to the production configuration, equipment, plant locations or process technology, the emphasis should be on making steady improvements to the process operations. Doing so is often sufficient to mitigate the generation of Waste which would allow the Product Family to be manufactured efficiently and at a low cost, thereby increasing customer satisfaction and improving profitability. However, as the implementation of the Future State Design is time-critical, the incremental process improvements need to be achieved in quick-time, an aspect which necessitates that the manufacturing team collaborates in mission-mode. This rapid and time-bound approach to Process Improvement is known as Kaizen Burst or Continuous Improvement Blitz. Kaizen Burst is a hybrid of the two Japanese philosophies - Kaizen and Kaikaku.
Processes earmarked for Kaizen are to be incrementally improved over a long time-horizon. For example, if Mapro seeks to reduce the energy efficiency of its Boiling process by 5% in a year, then a small improvement of ~0.4% can be targeted every month. This can be accomplished by adjusting temperature controls, performing equipment maintenance, altering material composition, and so on.
On the other hand, processes earmarked for Kaikaku are to be radically improved over a short time-horizon. For example, if Fenesta seeks to increase the output of its uPVC Profile Extrusion process by 40% in order to meet the anticipated surge in customer demand next month, it can set up a task force to find viable solution(s) on high priority - some of which could be to procure/lease additional extrusion equipment, increase duration of a production shift, adding a production shift or outsourcing production to contract manufacturers.
A Kaizen Burst or Continuous Improvement Blitz is a hybrid of these two philosophies - while the improvement target is moderate (typically <20%) and the mode of improvement is incremental in nature (akin to Kaizen), the timeframe of achieving it is short (akin to Kaikaku). A Kaizen Burst event typically lasts for just a week. The priority of the mission necessitates intense collaboration between departments and regular assessment of performance vis-à-vis the key performance indicators (KPIs). Attractive incentives are also usually on offer, in a bid to engage the workforce. Some of the ways in which the improvement target can be achieved through moderate modifications, albeit within a short time span are - by retooling an equipment so that it can change over into producing another product in quick-time, by providing training so that the operators can produce more reliably and with fewer defects, by redesigning the workstation to boost operator efficiency, and by maintaining the equipment so that it is less susceptible to breakdowns.
In a Value Stream Mapping context, the idea is to spot the source(s) of Waste from the Current State Map and subsequently develop a Future State Design which generates very little of it. Thereafter, the processes that would need a boost in efficiency in order to seamlessly transition to Lean Manufacturing have to be identified. This top-down approach of identifying the scope for having targeted Kaizen Burst events using a Value Stream Map is a much better alternative to the traditional Process Improvement initiatives undertaken in an ad-hoc manner based on anecdotal observations.
4.7.5 Continued
There is a need to improve both the manufacturing processes of Acme's Future State Value Stream - Weld+Assembly as well as Stamping. The former Continuous Flow workstation is the Pacemaker and hence, improving its efficiency is vital - we already know that Assembly 1 is a bottleneck as it churns out output 2 seconds slower than TAKT. Additionally, Acme also has to find a way to accommodate 30 Welding Changeovers in a workday. Let me elaborate the Kaizen Burst events that Acme has decided to undertake to ensure that its Future State Value Stream functions reliably-
KAIZEN BURST 1 - IMPROVE UPTIME OF EQUIPMENT AT SPOT WELD 2
As evident from Acme's Current State Map (Figure 48), the Uptime of Spot Weld 2 stands at 80%. This means that its equipment was out-of-order for 20% of the available working time, historically.
This calls for a Kaizen Burst event to be scheduled, with the objective being to minimize Equipment Downtime as all the operations within a Continuous Flow workstation need to be extremely reliable and synchronize perfectly with each other.
Acme has determined that by utilizing better equipment maintenance techniques during the Kaizen Burst event, the downtime affecting Spot Weld 2 can be eliminated completely.
KAIZEN BURST 2 - IMPROVE CHANGEOVER TIME OF SPOT WELDING OPERATIONS
Minimizing the Changeover Time of both the Spot Welding activities from the existing 10 minutes will be the objective of Acme's second Kaizen Burst event. This time has got nothing to do with setting up the Welding equipment (Weld Gun) to perform a basic fabrication procedure, rather it has to be related to the time elapsed waiting to receive the Stamped Brackets from the Stamping process.
The Supermarket Pull System devised for Raw Material as well as for Stamped Brackets would ensure a steady supply of Stamped Brackets to Spot Welder 1, thereby helping him to wait less in the future. Acme has also decided that the material handler would load the Stamped Brackets Bin brought from the Supermarket directly onto Gravity-feed Racks at the workstation. This will allow Spot Welder 1 to have ready access to his inputs, enabling him to save some more time - recall that unnecessary Motion is a Waste, and this initiative demonstrates the utility of mitigating it.
Additionally, as the Material Flow within a Continuous Flow workstation is uninterrupted, Spot Welder 2 will stand to receive an instantaneous supply of Welded Brackets from Spot Welder 1 which would help the former to completely eliminate his time spent on performing a Changeover.
All in all, Acme has determined that it can incrementally reduce the Welding Changeover time to just 40 seconds (~93% reduction) through this Kaizen Burst event. Thus, the total time to perform 30 Changeovers will be 1200 seconds or 20 minutes in a workday. The time remaining for Production would be 55200 seconds − 1200 seconds = 54000 seconds or 15 hours which implies that the Continuous Flow workstation would need to produce 920 units of Finished Brackets at a Cycle Time of ~58 seconds i.e. just 2 seconds quicker than TAKT, which is highly preferable.
KAIZEN BURST 3 - IMPROVE CHANGEOVER TIME OF STAMPING OPERATIONS
Moving upstream, the 200-Tonne Stamping Press has a long Changeover Time of 1 hour. Reducing it would not only improve the efficiency of the Stamping process, but also have a positive ripple effect on the Pacemaker process as well, as indicated in Kanban Burst 2.
Acme is confident that it can reduce the Changeover Time to 10 minutes (an 83% reduction) through this Kaizen Burst event as the way to reduce the setup time of the Stamping Press equipment by retooling it is well-known in the industry. Moreover, the Supermarket Pull System devised for Raw Material will also help ensure rapid, on-demand transfer of Steel Coils to Stamping which will also help in reducing the Changeover Time.
KAIZEN BURST 4 - IMPROVE OPERATOR UTILIZATION AT WELD+ASSEMBLY
The first and second Kaizen Bursts would target improvements at the Pacemaker process of the Lean Value Stream - completely eliminating the downtime of Welding equipment and reducing the Welding Changeover Time to well under a minute, respectively. I had taken a detour for the third Kaizen Burst, which would target an improvement at the Stamping process - to reduce its Changeover Time to 10 minutes, as this will also be one of the contributing factors to the success of this fourth and final Kaizen Burst which would target to improve Operator utilization at the Pacemaker process. From the Cycle Time information of the four operations captured in the Current State Map (refer Figure 51 below), it is evident that the total work duration at this Continuous Flow workstation would be 187 seconds, which translates to an average of ~47 seconds per operator. This is a far cry from TAKT and indicates that the operator resources at this workstation will be significantly underutilized.
If Acme were to run the Weld+Assembly Continuous Flow workstation with one operator less i.e. with a total of 3 operators, then the average work duration per operator would become ~63 seconds (187 ÷ 3) which is still above TAKT but can likely be improved to match it. For this new arrangement to function though, Spot Welder 1 would need to operate a portion of Spot Weld 2 as well, Spot Welder 2 would need to operate a portion of Assembly 1 as well, and Assembler 2 would need to operate a portion of Assembly 1 as well - this shuffling of work elements would require Acme to facilitate on-the-job-training for the three operators so that they can perform their new activity with ease besides their original one. The success of this endeavour would mean that Assembler 2 can be relieved from his current duties and be deployed productively elsewhere.
Acme has assessed that due to the positive repercussions from the previous Kaizen Bursts as well as due to the uninterrupted Material Flow occurring within the Continuous Flow workstation, the total work duration within Weld+Assembly would reduce to 168 seconds (10% reduction). Running this workstation with 3 operators will be viable now as the average work duration is 56 seconds per operator - better than the 58 seconds target derived in Kaizen Burst 2 after reducing the time spent on Changeovers from TAKT. Assembly 1 would also cease to be a bottleneck.
Thus, an action plan to address the pressing issues has been devised. While the Kaizen Burst events have been scheduled keeping in mind the time-criticality of Future State implementation, incremental Process Improvements using Kaizen would continue even after. After all, Continuous Improvement is to be pursued for perpetuity. Acme is now ready to begin its transition to Lean Manufacturing!
4.8 SUMMARY, TWI INDUSTRIES CASE, CONCLUSION AND ABOUT US
Below is the summary of the changes that will be made to Acme's Current State Value Stream of Steel Brackets, utilizing the Lean Manufacturing guidelines-
All in all, while obtaining requisite buy-ins for the Future State Design from all the stakeholders involved and especially the robust implementation of it would be key, Acme has already laid the foundations for a seamless transition to a low-on-Waste and high-on-Performance Value Stream. If everything goes to plan, Acme will be able to reduce the Production Lead Time for its Steel Brackets Product Family to a shade over 5 days which is a staggering improvement of ~79% from the 23.6 days right now! The Customer Lead Time component of it would shrink even further, which is the time the customer has to wait to receive the product after placing a Purchase Order, to just 2 days thanks to the installation of the Finished Goods Supermarket. This means that Acme will reap the benefits from having a high Inventory Turnover rate. That this boost in efficiency has been envisaged by minimizing Waste and by fine-tuning the Material and Information Flow alone and not through radical initiatives such as procuring advanced equipment or outsourcing manufacturing reflects the utility of pursuing Lean Manufacturing, facilitated through the incredible Value Stream Mapping technique, in the ultra-competitive industrial world of today.
IN THE MOOD FOR MORE?
I had initially intended to cover another interesting case pertaining to the Steering Arms Product Family at TWI Industries within this post itself. This case is a contrast to Acme in terms of the characteristics of the Material Flow and how Lean Manufacturing is incorporated within - the readers would enhance their grip on the concept of Value Stream Mapping. However, as I have breached the word limit set by my website service provider (Overproduction😅!), you may refer to the video demonstration of this case from this timestamp.
As a parting suggestion, in case you have liked what you've read, do pursue Value Stream Mapping yourself - be it for your manufacturing operations, service operations or even your daily personal routine. While the core characteristics remain the same, the applications of this concept are numerous.
ABOUT US
Intelloc Mapping Services, Kolkata | Mapmyops.com offers Mapping services that can be integrated with Operations Planning, Design and Audit workflows. These include but are not limited to Drone Services, Subsurface Mapping Services, Location Analytics & App Development, Supply Chain Services, Remote Sensing Services and Wastewater Treatment. The services can be rendered pan-India and will aid your organization to meet its stated objectives pertaining to Operational Excellence, Sustainability and Growth.
Broadly, the firm's area of expertise can be split into two categories - Geographic Mapping and Operations Mapping. The infographic below highlights our capabilities-
Our Mapping for Operations-themed workflow demonstrations can be accessed from the firm's Website / YouTube Channel and an overview can be obtained from this brochure. Happy to address queries and respond to documented requirements. Custom Demonstration, Training & Trials are facilitated only on a paid-basis. Looking forward to being of service.
Regards,
