Every Monday, LogistiekProfs.nl lets a logistics expert do the talking. This week, Vincent Wiers, expert planning & scheduling, asks whether major decisions in logistics should be made locally or centrally. Should we leave autonomy to or rather take it away from the shop floor?
Decentralisation is used by many companies to make better decisions and operate more decisively. The idea is that decisions are made wherever the information is available to make the decision. This is especially true for planning and scheduling. One might ask whether there is any point in sending a detail schedule to the shop floor if the shop floor can just as well decide for itself what work is best to be done next. There seems to be a contradiction in this decision between the more hierarchically oriented concepts and the sociotechnical or lean concepts. A true business answer to the question 'central or decentralised' is: it depends. It can sometimes work very well to make decisions locally, and it also saves expensive planning capacity and systems.
Read also: The job requirements of a good logistics planner
Local: lack of insight across the chain
In some cases, however, it does not work well and autonomy is better taken away from the shop floor, however unsympathetic this may seem. Suppose there is a series of departments supplying each other, and one of those departments is the bottleneck. And now suppose that bottleneck can only be taken care of properly if the upstream departments loosen the efficiency reins a bit. When decisions are made locally, chances are that the own department sets the priorities, and the bottleneck suffers. A lack of understanding about the whole chain is the most obvious reason for not granting local decision-making power. Another reason may be that planning and scheduling are very complex and it is not clear to someone on the shop floor exactly what consequences are attached to a particular order of work.
Complex situations
Complexity in execution is thus a key feature in granting autonomy. Another characteristic is the need to make decisions locally - the amount of disruptions. If the situation is not too complex but there are many disruptions (some job shops meet this), the obvious thing to do is to schedule locally and not send detailed instructions to the shop floor. At the other end of the spectrum, we have situations that do have complexity and low uncertainty, such as some (semi) batch processes in, say, a chemical production system. But what to do with high uncertainty as well as high complexity? It is not easy to say, but it could be a combination of central scheduling and re-scheduling, and reducing disruptions through process improvements. The most important thing is to make a situation manageable, because without predictability, planning is impossible.
Vincent Wiers - Expert planning & scheduling
