Future of Lean: is robotic motion/transportation waste? | Christian Hohmann

Michel Baudin‘s comments: It’s a valid question, but one that should be asked about handling and transportation automation in general, not just robots. It is also one that is not properly answered with the simplistic theory of value and waste that has been reiterated in the English-language literature on Lean for 20 years.

Robots are attention grabbers. On a factory floor, they are highly visible, and their motions are fascinating to watch but it shouldn’t distract us from other devices like Automatic Storage and Retrieval Systems (AS/RS), Automatic Guided Vehicles (AGVs), or self-driving trucks, and the control and sensing technology that enables these devices.

As pointed out in earlier posts on muda, not all motion and transportation is waste, only unnecessary motion and unnecessary transportation.  It is, strictly speaking, the meaning of the original Japanese list of 7 waste categories, and the exact translation of muda is unnecessary.  It is absurd to consider that moving a screw from a bin to a hole is waste; on the other hand, walking back and forth 6 feet to do it is. Transporting finished goods from Final Assembly to Shipping is not waste, but taking WIP back and forth to a warehouse between every two operations is.

Clearly, whether a task is done by a human or a machine, unnecessary means unnecessary. Let’s say you are using a gantry robot to pick documents from 6 bins and place them in a box containing an electronic product. It is a necessary operation. If you place the 3 document bins on each side of the box instead of all 6 on the same side, you cut in half the robot’s travel distance, which may reduce the total operation time by, say, 20%. If the document packs are custom for each piece, you may further reduce the robot’s travel time by locating the most frequently picked documents closest to the box.

While motion is from bin to workpiece or from one station to the next inside a line, transportation is between warehouses and lines or between lines inside a plant, as well as between plants, consolidation and distribution centers, and customers within a supply chain. Within a plant, the main issue is the number of transportation steps rather than their lengths.

Within a plant, the main issue is the number of transportation steps rather than their lengths. Each step requires materials preparation, loading, and unloading, all of which consume more resources than the actual movement from A to B. It does not make much difference whether A and B are 50 or 500ft apart. Improvements come from eliminating transportation steps rather than shortening them, and it involves collocating machines and integrating them into flow lines. It’s not about the technology you use for in-plant transportation. If, however, you reduce the number of transportation steps inside the plant by 80%, it has an impact on the economic justification of automated transportation.

In the supply chain, on the other hand, it’s distance that matters but improvements come more from the way you structure it than from the level of automation you use. Self-driving trucks will not do much to improve performance in a badly designed supply chain. If you rely on 3rd parties for transportation, then the choice is theirs. If you use your own trucks and drivers, you may use the drivers to validate the shipments in some fashion when they pick them up at suppliers, which won’t be done by self-driving trucks.

Strictly speaking, in everyday life as well as in production, an activity is waste if and only if nothing gets worse when you stop doing it. In production, if you stop doing X and quality does not suffer, costs don’t go up, deliveries are not affected, nobody gets hurt, and morale is not degraded, then X is waste. This is observable. You can measure it, and backtrack if you realize that X was actually useful.

The theory of value based on customers willingness to pay sounds good until you try to unpack its logic. Sales reps know what customers are willing to pay but, in a shopfloor situation, you are many layers away from these customers. But what about “internal customers?” This means the next operation. “The next operation is your customer” is a metaphor from Total Quality Control that does not apply here. The next operation doesn’t pay you and therefore you have no way to assess its willingness to pay.

Since willingness to pay is not applicable here, the fallback is to declare that only activities that physically transform the product add value. Everything else, from posting up-to-date instructions to delivering parts to the line is “waste.” The many such activities that are necessary are then called “necessary waste,” an oxymoron when you consider that “waste” here is used as the translation of the Japanese word for “unnecessary.”  What is the effect of branding necessary support activities as waste? Offending and demoralizing the employees who carry them out.

The developers of TPS did not back it up with a good theory. As Georges Matheron put it “nothing is as practical as a good theory.” It would be particularly useful when deploying something like TPS outside of its original context. Developing a good theory, however, is an intellectual challenge. What the Lean community has come up with so far is not it.

Here are a few earlier posts on these topics:

• Occam’s Razor, Value Added, and Waste

• Absence of “Value Added” in the TPS literature

• Management Whack a Mole and the Value of Lean