Recently, I had the following straightforward conversation with a client that was thinking about adding more technology into their manufacturing and supply chain mix. In this article, we are picking up part way into the conversation—and some minor details may be changed to protect the privacy of our client.
There is a lot to be considered here. I will try to address a few matters in a coherent way, although there is no way I will be able to answer every possible question that might arise in these emails. (In fact, initially anyway, it is possible that my responses to some matters may actually raise more questions in your mind than they resolve.)
You said: “Costing and time considerations are presently very abstract and not real-world” for you
I wouldn’t be too concerned about that, if I were you. In my opinion, and in the opinion of some other highly respected individuals, much of the needless complexity we find in the business world—and the manufacturing world, especially—is cause by a single misconception. This singular misconception is, unfortunately, a deeply held belief. By that, I mean that this concept—this belief—is held at such a deep level in our being that we almost never consciously challenge it.
That deeply held belief could be stated in any number of ways, but for the sake of this conversation let me put it this way:
Because financial information is essential in defining the purpose of a for-profit enterprise, we hold deeply the belief that financial information will also provide us with the primary means to control the outcomes or results of the enterprise (or, supply chain).
As H. Thomas Johnson, Professor of Business Administration at Portland State University, stated so cogently in his article “Managing a Living System, Not a Ledger” (Manufacturing Engineering, Society of Manufacturing Engineers (SME), December 2006), “Until that idea is abandoned and the practices it spawns cease, there is no reason to believe that ‘lean’ in any of its forms… can improve the long-run performance of businesses or the economy as a whole. This becomes evident when one examines why companies seem unable to match Toyota’s financial performance.” [Regarding Toyota’s financial performance when compared to U.S. automakers, read here.]
If your company’s (or, supply chain’s) goal is to MAKE MORE MONEY, then FLOW must be the objective of all that you do.
Toyota’s operations are very efficient; but they never focus on—or even measure—efficiencies. This efficiency is brought about, not by focusing on unit cost, batch sizes, and individual employee performance. Rather, the inherent efficiency of the Toyota Production System is the result of an irrepressible focus on “continuous improvement through endless rapid problem solving,” and FLOW [Johnson – ibid].
“[T]he Toyota accounting system treats daily plant operations essentially as a ‘black box’ that it does not enter. Accountants, of course, record everything that goes into the plant, and all the production that comes out.” But, the finance department at Toyota never concerns themselves with “costing” of operations or using financial metrics as “levers” for improvement. [Johnson – ibid]
Is there a place for “ongoing measurement of time usage”? Yes.
Time identified as “waste” should be, inasmuch as possible, be eliminated. But this is done by improving the processes and FLOW of manufacturing, not by identifying waste and accounting for it on a case-by-case, work order-by-work order basis. This leads to too much additional non-value-added activity. (The effort to capture and account for the wasted time is to add one non-value-added activity on top of another. This is the opposite direction from where you want to go.)
Is there a place for “time studies”?
Of course. Time studies are valuable for use in forecasting capacity requirements based on change in products or processes.
If you are going to keep a bill of materials (routing) similar to a typical product we reviewed when I was last on-site with you and your team, and you intend to schedule every job at each work center, capturing the starts and completions on each of the operations, then your non-value-added time will necessarily skyrocket. Consider the following:
With a typical routing of eight (8) operations and processing 250 orders per day, you will need 2,000 data entry operations just to capture completion times—double that if you want to capture both start and stop times. If each entry takes an average of 20 seconds to complete—end-to-end—you will lose more than eleven (11) hours (1.4 person-days) per day to the data entry burden. All of this adds nothing to the value of the products going out the door.
Take a good look at the first block of “requirements” that you sent me. These are really prerequisites more than requirements. These are prerequisites that the newly proposed technology will require of you and your team in order to provide the “results” they are predicting in the post-implementation world.
Note the repeated use of the term “accurate.”
- Identify and integrate accurate work center capacities
- Implying accurate cycle times
- Implying accurate set up times
- Implying accurate work center allocations
By the way, for a routing with eight (8) operations, that is at least eight (8) cycle times that must be accurate, because scheduling errors will compound as they run through the routing operations. Then, of course, there are the set-up times and work center allocations that also must be accurate, in order for the schedule produced by the APS (Advanced Planning and Scheduling system) to approach anything near reality—even in theory.
The assumption made by every APS is “accuracy.”
For better or for worse, the APS will schedule according to the data you provide, and it will do so “accurately”—that is, in theoretical compliance with all the data you provided to the system.
However, no matter how accurate you attempt to make all of these inputs, the schedule produced by your APS will NEVER represent what is going to actually happen on the shop floor. This is because the “clairvoyance” component of the APS system is still under development.
Your APS system cannot know that Betty strained her back last night is only working at 72.3 percent of ‘standard’ capacity right now, and that number is falling throughout the day. Nor can your APS know that Joel had a fight with his wife before coming to work and is working at only 81.9 percent of ‘standard’ capacity; or that Mary Jo is feeling really good about herself today because she knows she’s “going gangbusters” (and working at 111.2 percent of ‘standard’ capacity). But all of that is going to change tomorrow, or later today.
In my opinion, an effective demand-driven scheduling system should schedule your CCR(s) [Capacity Constrained Resource(s)] only, and these should be protected by appropriately designed buffers of either stock, capacity or time (or some combination of these). It is your CCRs that limit your company’s ability to make more money.
In a high-volume MTO (make-to-order) environment like yours, where the CCRs may change due to the mix of orders being processed, then it may be valuable (or, even, necessary) to artificially “fix” a constraint to function as your “drum”—to set the pace for operations. This is why simpler is better; it simply required far less data and produces results that are far more accurate in terms of reflecting reality.
Murphy must be accounted for in the design of your buffers, because it is impossible for any APS system to account for Murphy (by definition, attacks of Murphy are random and cannot be built into any scheduling mechanism). As I said, your buffers may be any one of three types: stock, capacity or time.
With a simple approach to scheduling, you will know if your system is functioning on-time and according to the anticipated schedule simply by knowing the condition of all the buffers in your system.
If your buffers are green or yellow, you are in good shape. If any buffer is red, you know immediately know where to focus your attention and, based on buffer penetration, you will know how to prioritize your actions.
That is inherent simplicity at work. You will see some of that in the webinar I recommended to you, I am sure.
I know this is a lot to swallow all at once. And there are probably about 20 more pages I could (attempt) to write to try to convey more depth about some of these concepts. But, it’s too much to attempt now.
I hope this isn’t so poorly written as to leave you more confused than helped. I was just trying to address a couple of points that really stuck out to me.
Let me know if I’ve left you worse off, instead of better, for having had this conversation this morning.
We hope you found this dialogue interesting. If it struck a chord in something you have experienced or to which you can relate, please let us know. Leave your comments below, or feel free to contact us directly. Thanks.