Note: For Tools in the A – L Range, click here.
A brief article discussing the benefits and methodology of making the transition to Make to Order is explained here.
The concept is more fully explained in our article “From Make to Stock to Make to Order.”
As an organization moves toward “Lean”, many of the traditional measurement devices fail to perform accurately or timely enough. People are cross trained and move back and forth to various jobs. Operators perform minor maintenance. Natural work teams work on improvement projects. Kanbans will cause short term delays, etc. Measurement often must move from individual performance to team performance. In some progressive companies, the teams themselves provide input into the individual reviews. There is, however, no substitute for being “out there” on the shop floor.
The implications of making such a transition are far from trivial. A transition to Lean Manufacturing can generate a heap of up-front tax-free cash, but adversely affect your short term accounting profits, require an overhaul of your measurement and reward systems, change your typical work week, and modify your perspective in regard to make/buy.
The rewards of such a transition, if done correctly, can be huge. As Nike would say, “Just Do It”!
In a lean environment, we are always looking for a better way to do things. Seek a performance measure that encourages process method improvement. This is best accomplished by comparing current “actual” cost to historical actual cost. If improvement is what you seek, it is pretty much irrelevant what it “should cost” based on some theoretical “standard”. What really matters is “is it currently costing me less than last year? Last month? Last week?
Instead of spending time and energy calculating a theoretical “standard,” you will be much further ahead if you focus on finding a better way, and on providing people timely feedback as to how well they are performing. If you feel compelled to have a performance against standard system, at minimum calculate performance vs. both the current standard and against a fixed standard. The former measures theoretical level of effort (current standard). The latter measures their impact on profitability and rewards people for innovation (methods changes) as well as effort.
Unit optimization has another insidious impact. It adds to complexity.
We’ve been working with metal producers since 1989. As a result, a different approach and a few unique techniques have been developed to cope with these issues. The results have been dramatic. To date, we have worked with seven different metal producers, world wide.
Our industry average return on investment exceeds 90 to 1.
Milk Runs allow for small, frequent, regular material deliveries / pickups.
We worked with a steel pipe mill. Maintenance had been neglected for years, and the line was regularly shut down as one after another problem occurred. An interim solution was to have each operation develop a list of 2 hr maintenance / reliability projects. The rule was established: When any piece of equipment breaks down, the entire line will remain shut-down for at least 2 hrs while everyone works on their area projects. Within a few months, total line up-time was significantly improved.
The idea is to re-engineer the maintenance processes to allow them to be done in smaller increments, “modules”. Thus, instead of scheduling one 24-hour maintenance period to be done every month, we would seek to replace this schedule with one 6-hour down period, each week. (See “Total Productive maintenance” and “Cascade Planned Maintenance”)
By having access to the idle equipment on weekends, an excellent Preventative Maintenance program was implemented, and equipment upgrades were addressed in a timely manner. Within six months of the change, this issue was no longer a problem and productivity jumped by 20%!
Net result: Where our company saw a 10% change in demand, our suppliers will see a 20% change in demand! Thus the “Multiplier”. The same effect happens when demand increases. And, the further down the supply chain, the greater the swings.
You can readily see that the extent of this multiplier is proportional to the amount of inventory in the supply chain, and the number of levels in the chain.
What can you do about it? Reduce inventory throughout the supply chain.
Note: your customers will carry inventory of your product in proportion to your lead time and your delivery reliability. Do these things well, and you’ll minimize the multiplier. You’ll also want to monitor and compare changes in your product’s demand to changes in top level demand.
Analysis of the recession of the 1970’s showed that while the end demand for steel products fell ~ 10%, the demand on the steel industry (the bottom of the supply chain) fell by over 40%! Needless to say, a “Lean” supply chain is essential. Maintaining short lead times, and reliable deliveries to your customers is crucial (see “Boom – Bust Cycle”)
Note: We are NOT forming teams for teams sake. We are squeezing out the inventory, which results in the formation of “teams” whether you want them or not! It is at this time, when people are closely interdependent, that additional teaming skills training will be needed. Books like “Zapp” and “Heroz” can be useful training aids. (See “Training: Do a Chapter a Week” and “Problem Idea Charts”)
While this discretionary budget can be both an employee motivator and a practical process improvement aid, this is one area that we would NOT advocate “just do it”. Spend some time. Think it through. Establish clearly defined “acceptable uses” for the spending. Create a “straw man” procedure as to how the team would make spending decisions. Discuss safety precautions that the teams must consider.
Some companies have established team accounts at places like Grangers, and allow their teams to place orders against these accounts. Others provide debit cards. Remember: It is a whole lot easier to increase the spending limits than it is to decrease. Start slowly, and then build upon it. And don’t second-guess valid attempts that did not work out. These budgets are typically small, and therefore low risk. It is not at all unusual for the teams to spend several hundred dollars on items that had little or no positive effect, before they eventually hit the project that saved the company tens of thousands of dollars. The idea is to encourage sensible trial and error!
It has been our experience that NPI has an 80/20 law of its own. Twenty percent of the time and cost will be required to develop a viable concept. The remaining eighty percent will be required to bring this concept to a saleable product. Note that in almost every industry, “time to market” continues to grow in importance. It IS possible to set deadlines. And it is critical to set a cut-off on design changes (for the introductory model). In today’s environment, it is almost always preferable to have “good” product out first, rather than a “great” product out late.
Needless to say, if we are ever to drive out this resulting finished goods inventory and associated hidden waste, we must measure, set goal curves, and improve performance on on-time completions. Measure the percentage of completions that occur exactly on the date scheduled. Then set goal curves to dramatically improve this performance. On-time completions should become a critical gauge of “company well-being”, and treated accordingly. Establish a mantra of “on-time, all-the-time”
Beware: Some companies delude themselves by saying “we talked to the customer, and he said it was OK.” In actuality, all we’ve really done is reinforced the idea that the customer should ask for his product earlier than he really thinks he’ll need it. Treat on-time delivery, as measured against the original promise date, as another critical quality parameter.
Note: Short lead times are of little value if your delivery performance to your promise is not near perfect. However, short lead times combined with reliable delivery performance can be a significant competitive advantage. Caveat: Prove out that you can perform BEFORE you advertise the capability.
One of the most difficult obstacles that many companies must overcome is the fact that their current measurement and reward systems target “unit optimization”. The easiest way to think about this is to picture an automobile assembly line. Every piece of equipment is paced to the rate of the line, yet most of the equipment could produce considerably more than that pace. Doing so, however, would require queues of inventory between operations, different crewing schedules, etc. We could indeed increase the effectiveness of the individual pieces of equipment, but how efficient do you think the total factory would be?
Note: There are techniques that can and should be applied to individual units that have been sub-optimized to improve their effectiveness as well. However, the biggest gains come from the initial phase of optimizing the whole. Take a look at the cost of complexity that unit optimization brings about.
The typical example is the taking of an excellent technical person (engineer, software programmer, etc.) and promoting them into a management position that requires an entirely different set of skills. The problem is exasperated by the unwillingness, in many organizations, to correct the situation when an individual is clearly struggling.
Many innovative companies now provide a corresponding technical path that similarly rewards key individual contributors. This allows them to continue doing the things that they are equipped to do well, while achieving the financial and status rewards similar to the progression available through the management ranks.
We are continually amazed at the number of employees that have no idea what is being done in the other areas of the factory. Take the time to provide a tour for your operators. You may be surprised at how it increases there caring and awareness of opportunities for improvement. Note: This same concept has produced huge benefits when extended to customer and supplier sites.
Wouldn’t it make more sense to put the parts immediately where they can be used for production? Now, I understand that lots of groundwork must be done to accomplish this logical step. Just like the Frito Lay delivery guy, let’s get the supplier, where ever practical, to deliver quality parts (we’ve pre-qualified his process), in reasonable batch sizes (we’ve shown him how to cut his set-up times), directly to the point of use. Transaction costs can be minimized through backflush.
Note: Many tool supply companies will now furnish “vending machines” that are placed throughout the shop floor and are stocked with commonly used tools and supplies, generally on consignment. Operators/supervisors are provided with credit-card type keys that record when, what, and by whom, each part has been used.
The pyramid is typical of electronic assemblies where a large number of component parts are used to make up a few products.
The Hourglass configuration is found where a large number of component parts make up a limited number of subassemblies. These subassemblies can be configured into a wide variety of finished products (flavors).
Push systems have a tendency to constipate the system, adding to WIP and necessitating expediting. (See “Kanban Controls”)
While the idea of PPV is quite appealing, it has a major shortcoming. It ignores all of the other costs associated with a purchased item: Transaction costs, Freight, De-trashing (getting rid of packaging material), Receiving, Incoming inspection, Stocking, Record keeping, Pulling the material from stock, Shrinkage etc.
A higher priced product that is vendor managed, delivers directly and frequently to the point of use, just in time, with perfect quality, and re-usable packaging, etc. may be a far better bargain.
In most cases, we suggest replacing PPV with a measurement that includes the “Total Cost of Purchase”.
Establish turn-around targets for “time to disposition”. Measure it. Push to assure that all functions required to make such disposition are available as close to 24 – 7 as feasible. You can’t have your operations waiting on material that may, or may not, be available for production.
One effective technique is to require that all discrepant material remains within the kanban allotment, i.e. if you are allowed 20 pieces between operations, and 5 are on hold (discrepant), then only 15 “good” pieces are allowed in the kanban. This will add operating pressure for quick discrepancy resolution.
A cursory study of each product’s historical demand, cost of processing, and cost of carrying the additional inventory can lead to simple decision rules. The various products are coded. Green: Cut up the entire coil. Red: Cut only what is needed, then put the unused portion of the coil back into stock. Yellow: Use your discretion.
Also see SMED: Single Minute Exchange of Dies”
I suggested that they rent some space at their primary customer’s site (this customer used 80+% of their volume), cut a hole in the wall, and deliver bottles directly to the bottling line.
The result: No pallets. No shrink-wrap. No storage. No transportation. And a customer for life.
Note: while lead times may float for the B’s & C’s, reliability cannot. Credibility is paramount. Say what you’ll do, then do what you say.
What SHOULD you expect from a lean transition? Large, tangible corporate-wide results.
If you are dissatisfied with your lean progress, give us a call. It won’t cost you a dime, and I guarantee that you will find the discussion “Value Adding”.
If we begin to drain some of the water from our pond, some of these rocks will be exposed. We will then have to remove them so that our ship can continue to sail. The same occurs in our factories as we begin to reduce the inventory. Lot size / set-up time, vendor reliability, lack of flexibility, un-reliable processes, etc. will slowly be exposed as the inventory is reduced. Fixing these “rocks” reduces cost and improves quality and customer service.
Note: A company-wide inventory reduction process serves to “optimize the Whole plant.”
See the analogy below.
Nature abhors a vacuum. Take aggressive steps to preserve the empty space or it will rapidly refill!
Rough Cut CRP uses the Master Production Schedule (MPS) to calculate labor requirements. The typical rough cut CRP algorithm uses the MRP / ERP system master schedule plus a rolled-up bill of labor to calculate future labor requirements.
As an example, the MPS calls for us to build 20 men’s 26” bicycles in period four. The labor requirements for period four would be 20 times the rolled up bill of labor (which would include all fabrication, subassembly, final assembly, test, and pack labor requirements for a men’s 26” bike.)
To be reasonably accurate, Rough Cut CRP requires flat bills of material, short lead times, small lot sizes, and little or no safety stock requirements, and works best in “make to order” environments. It is generally adequately accurate in a lean organization. (See Capacity Requirements Planning, CRP). !
Our article, The Rubber Factory, details this powerful technique. (See “Counter Cyclic Products”)
A salary Increase Matrix is easy to explain to your employees and is generally seen as fair.
The management team agrees on the matrix parameters, i.e. how to pay your people based upon their job performance and their current level of pay. Then the emphasis changes to accurately and consistently evaluating the performance of your people.
As an example, Slim is an outstanding employee (A performer) and is currently being paid in the first quartile of his salary range. The matrix suggests that Slim receive a substantial increase (10-15%) in his pay. The logic is that an outstanding performer should be paid in the upper part of the salary range.
Similarly, Mark is an average performer (C rating) but is being paid in the top quartile of his job pay range. The matrix suggests that he is currently being overpaid for his contribution, and his salary should be adjusted accordingly.
You will soon, however, run into a problem. The typical multi-shift operation has shifts that are “back to back”, i.e. one shift ends at the same time that the next shift begins. If your process allows, it can be a significant benefit to stagger the shifts, thereby leaving a gap between shifts for recovery, repairs, and preventative maintenance; e.g. 1st shift runs 7:00 AM – 3:30 PM, 2nd shift runs 7:00 PM – 3:30 AM (3.5 hr gap). Each shift can then be held accountable to make the schedule, exactly, before going home. Instill a discipline: “ The shift ends when the schedule is done”. (See “On-Time Completions”)
A quick and easy fix was to apply a simple randomizer to the internal completion date. This smoothed the completions over the promise week, and provided producible schedules. The customer continued to receive a “week of” promise date, but internally the operating and scheduling people were measured to the completion day. Needless to say, the next obvious step is to begin providing daily ship date promises to the customer. So far, however, we haven’t been able to win that battle!
A similar problem arose at a steel finishing plant. Standard practice was for production scheduling to be done on weekdays, day shift only. However, production operations ran 24 x 7. Needless to say, as the internal lead times were reduced, it became unrealistic to attempt to schedule production that far in advance. The schedulers agreed on new work schedules. Schedulers were also provided with PC’s and the ability to access the scheduling software at home. This allowed for critical schedule changes to be made electronically from the schedulers home, if needed, 24 x 7.
Here’s an easy first step in establishing sequential inspection. Have every operator write down the items that they could possibly mess up. Pass this list to the next operator as his/her checklist. Run for a week or two, adding to and correcting the list. Then formalize (print up, add color photos, etc.) and post at each work-station.
The automobile assembly line and associated feeder lines are set up in this fashion. By “pulling” the correct parts in the correct sequence, a complete final assembly is produced with minimal WIP inventory, just in time. Note, that while this is by far the most efficient form of pull system, it requires a considerable amount of pre-work. Quality and reliability of all processes must be truly world class, as any defect or line disruption can throw off the sequencing. (See “Kanban”)
(See “5S / Area Organization”)
However, when lead times are excessive; lot sizes are too large; schedules are un-realistic (poor capacity planning and/or shop loading discipline); and accountability is lacking, SFC systems will be seen as a necessary solution. NOTE: They will NOT solve the problem.
In most industries, if Lean is done correctly, the need for SFC is greatly reduced or eliminated, and, in the few industries that truly do require SFC systems, their credibility will be greatly enhanced by a lean initiative. The “ cost of complexity” caused by these underlying drivers can be substantial.
What happens to WIP inventory levels and lead times if we were to move 2 of our operators to the 2nd shift? Operation 1 is done on day one, 1st shift. However, operation 2 now can also be performed on day one, on the 2nd shift. Same with operations 3 & 4. The lead-time and WIP inventory levels are cut in half.
Note: we did NOT increase capacity. Four operators still produce 1 completed unit each day. However, by manning more of the available hours, we are able to keep the product flowing, thereby cutting inventory, space requirements, and lead times accordingly. We used 8 hour operations for illustration purposes. Moving a portion of your capacity to the off shifts will have the same impact regardless of the work content.
For more information on this powerful technique, check out Making the Case for Multiple Shifts.
Many constraints are perceived, habit, or hear say. Require the letter of the law be shown. Don’t be afraid to challenge convention. All too often, it’s being done that way because “it’s always been done that way”!
Small machines often are simpler than the mega-machine and therefore more reliable and easier to fix. Mega-machines generally have a fixed output rate and crewing requirements. It’s either on or off. Multiple small machines, in contrast, can be crewed to provide variable rates of output. Note that doing so also keeps the labor costs in line with the actual volume required.
Note: If your operation is not already running 24 x 7, you may be able to attain many of these same benefits by increasing the available hours of operation. This concept is further explained in the articles “Making a Case for Muliple Shifts”, “Alternate Crewing Schedules” and “The Rubber factory”.
The typical process is to form a team involving the equipment operators, set-up people, maintenance, and an outsider or two. A standard changeover is then video taped. The team reviews the tape, identifying and challenging all elements. “Can it be done while the machine is running? Can it be done better, faster, cheaper? Do we need all off these screws/nuts? Can we use quick clamping devices?” Etc.
In areas where no previous set-up reduction efforts have been taken, it is quite typical to reduce the changeover times by 50%-75% with little capital outlay. (See “Quick Change Tooling”)
The problem is not that these endeavors are necessarily wrong. And it is certainly correct that people will learn from the effort. The point is that there is a more efficient and effective way of focusing the efforts.
If you wish to generate cash, reduce lead times, and prioritize your efforts in the most effective manor, use inventory / cycle time reduction to drive the process. Reducing inventory generates cash. It also highlights that next “rock” that must be attacked.
The Hands-On Group’s “Rapid Impact” process utilizes an inventory reduction goal, along with an on-time delivery goal, to drive the improvement process. As you progress toward these simultaneous objectives, problems will appear. Kanban installation, SMED initiatives, 5S blitzes, NWT’s, Vendor workshops, Cross training, Manufacturing cells, etc. are used as & when needed to fix these problems. The process of driving inventory out of the system acts to prioritize your specific actions. The specific “tools,” found in this article, are then used to achieve this inventory delta objective. And, the cash generated through the inventory reduction process will typically more than fund the cost of the transition.
Lean is about eliminating waste. Doesn’t it make sense to use a “lean” transition process as well?
Standardizing reduces procurement costs, complexity, and opportunity for error. Standardizing equipment reduces spare parts requirements and improves maintenance know-how. Standardizing processes aids in employee rotation and cross training/flexibility. And, the impact on quality is obvious. (See “Value Engineering”)
In our experience, a similar but much simpler procedure is often adequate: Divide the tolerance by 4. Consider the middle two bands around nominal as green. The outer bands are yellow. Outside of tolerance is red. Set up your SPC control charts and actions accordingly. (See “Six Sigma”)
The point is that it is not necessary to develop all the needed “tricks” by yourself. Take advantage of the wealth of knowledge already out there. Attend professional society meetings.
Send your people to workshops and seminars.
Participate in plant tours where available.
One distinct advantage of using experienced consultants is this transference of knowledge capability: The ability to “steal with pride” non-proprietary tools that can be similarly deployed at other companies and in other industries. Our 25 years of consulting experience at over 100 client sites has provided us a depth of understanding that is difficult to attain solely from books and seminars.
The advantages of commodity and part number methods include ease of training and ease of finding lost parts. The disadvantages are the poor use of space, and the difficulty in adding new part numbers. Random location systems maximize the use of the “cube” (warehouse space) but have no logical pattern and depend heavily on correct input of location data into the system. Most effective warehouses use combinations of the above. As an example, small parts might be placed in standardized bins in part number order. A card is then inserted in the bin identifying the location of any “over stock” that would not fit in the bin. Large or awkward parts are stored in random locations.
Walk the talk.
Caveat: As processes are linked to one-another via kanbans or other means, this task becomes considerably more difficult to enforce. Stopping one area, will soon lead to stopping upstream and downstream areas as well. There will be a temptation to exceed the kanbans and only work overtime in the affected area. Occasionally this may be justified. However, be very cautious as to the message being sent to the troops. Each exception makes the next exception a little easier, and can rapidly lead to a loss of discipline. (See “failsafe”) (See “One Perfect Unit Day”)
Phase two: Work with all key suppliers to reduce or eliminate the non-value adding activities typically associated with procurement: Product packaging, freight costs, receiving transactions, incoming inspection, stocking, accounts payable, record keeping, etc.
One of our clients posted all vendor performance charts right in the lobby for all to see.
For additional information, see our article “Lean Supplier Development.”
However, it has been our experience that this concept, while a worthy objective, is not very practical in many industries. With very few exceptions, customers do NOT buy linearly, i.e. the “takt time” is constantly changing. This problem has been alleviated in some industries (automotive, etc.) by artificially flattening “demand.” They accomplish this by establishing a production schedule for some fixed period of time (a week, a month). Needless to say, if actual demand is varying, and our production schedule is fixed, we must depend on finished goods inventory to take up the difference. This clearly violates the “PULL” concept of making something only when the customer requests it. As we shorten lead times and become more responsive to actual customer demand, the concept of takt time becomes less relevant. Instead, the emphasis needs to be put on flexibility and agility. (See “Rubber factory” and the post of the same name)
The point is: Don’t form teams just to form teams. Drive out the inventory, and natural work teams will automatically occur. Teach the teaming skills as/when the need arises. Not before. This issue and others are discussed in more depth in our article “How to Screw Up a Lean Transition.”
(Also see “Teachable Moment”)
When a salesman from the supplier called our client to say thanks, he was told “Wait ‘til you see the ad we have if you screw up!”
Remember: Companies don’t do things; people do. A little appreciation can go a long way in achieving sustainable world-class performance.
Like so many other good tools, this one is also subject to abuse. TOC becomes pretty trivial in environments that are not truly capacity constrained. The constraint, per the literature, is now lack of sales. Not much of a revelation! (See “Bottleneck”, “Optimize the Whole” and “Scheduling: Upstream & Downstream From a Bottleneck”).
We have been working with companies since 1988, and we’ve seen both extremes. We have seen companies abuse Theory Y: They provided no guidance as to what the company needs to achieve, no time frames for achievement, and no accountabilities. And, we’ve worked in companies that were so Theory X that people were afraid to make any decision, and would wait for specific direction for the most mundane of tasks.
Theory Y does not absolve management of their responsibility to set direction and provide clear objectives and accountabilities. And, Theory Y is situational: Responsibility must be doled out as the group / individual has grown to accept and handle it. In most organizations, the best balance has proven to be: “Management decides what needs to be done. Employees are provided considerable latitude in HOW those objectives are to be achieved.”
In general, Total Cost of Purchase is a far superior measurement than standard PPV (Purchase Price Variance). Lean Manufacturing focuses on reducing the Total Cost.
Multiple Lean techniques provide additional robustness to the entire quality process: Sequential inspection, failsafe devices, stop the line / fix the problem, etc.
The typical measurement is the unit cost (or labor hours per piece) required to get a product through each production unit. Management and operators are measured and rewarded for producing product through their department at the lowest processing cost. In the metal producing industries it is commonly referred to as “Tons per Hour.”
Using a “unit optimization” reward system is almost always an obstacle to overall company optimization and inhibits the successful transistion to a Lean Operating Environment. The reasons are fully explained in our post “How to Optimize your Entire Plant.” This is one of the measurements that you’ll likely want to get rid of.
Think of the engineering workload as a sine wave with peaks and valleys. Most companies staff the engineering department to handle the average workload. Backlog buffers the peaks and valleys so all engineers are pretty much constantly doing design engineering. If, however, you man at a level closer to the peak load, backlogs can be kept low (quicker turn around of designs) and the “idle time” can be used for value engineering, product enhancement, new product development, etc.
On a more narrow perspective, “value stream” is also used to denote the entire stream of activities required to move a product through the internal processes and activities within a company.
(See “Source Inspection”)
Here is a vendor checklist to help start the process.
Ideally, this room is dedicated to the lean improvement process so that the charts, notes, layouts, etc. can remain on the walls between meetings.
In practice, WIP is the designation of material that has been issued to the shop floor (often in the form of a kit). It is no longer considered WIP when it is completed and “closed” to finished goods.
From a Lean perspective, WIP is extremely important. It is the area over which we have the most control. The amount of Work in Process directly affects the time required to get a product through all manufacturing processes (lead time). It hides defects and delays their discovery. It adds complexity. And it hides scheduling and process problems.
Pressure to reduce WIP can be used to force a culture of continuous improvement..
Note: For Tools in the A – L Range, click here.
“As a company, we need to be faster in all kinds of ways–faster to decide, faster to choose, faster to act…. We need to take more risks, act on less perfect information.”
Worth Magazine 2000/02