For decades now, inventory has been considered a necessary evil. So yes, of course, inventory needs to be financed and impacts working capital.
The lean approach has contributed to the negative image of stock: excess stock is waste, a muda. And this excess stock, often linked to overproduction, itself a muda, can have an impact on many other mudas, such as unnecessary movements (handling), waiting times which result in longer lead times and less flexibility, and quality defects which are thus masked.
So yes, we need to eliminate waste, which means reducing unnecessary stocks. But beware, eliminating muda may involve first eliminating mura, another type of waste that concerns irregularity, but that's another topic...
Let's go back to inventory. While "unnecessary" stock should be eliminated, lean does not claim that all stock should be eliminated. For example, stock is essential if the time within which customers expect delivery is shorter than the time within which production or supply can be made.
Let's take the example of production: the absence of work-in-progress between production machines contributes to lowering the rate of production, i.e. the volume produced per hour. In the case of just-in-time production on an automated production line, if there is no buffer stock between certain machines, production stoppages on each of the machines immediately spread to all the machines and the effective production time falls. These production stoppages can be of different types, and their duration and frequency can be reduced, but they cannot be totally eliminated: stoppages for changing consumables or tools, adjustments, breakdowns and micro-stoppages.
The addition of buffers, i.e. places to receive work in progress between certain machines, makes it possible to decouple the operation of the machines and thus increase the time during which each of them produces. This leads to a paradox from the point of view of lean: adding stock allows for an increase in production capacity. Our DispoX solution allows us to carry out the relevant buffer sizing in the design phase with the Factory Design solution, and then in the operational phase to identify the bottleneck and evaluate solutions to remedy it, including the adjustment of certain buffers with the Factory Operations module.
Let's now take the more general case of the supply chain, stocks are omnipresent and their sizing is often poorly equipped: ERP and APS often have limited sizing possibilities. Even the Demand-Driven-MRP (DDMRP), which is very relevant for improving the management of operations, has only pragmatic and approximate solutions for sizing them. As a result, with these tools, stocks are positioned without any assurance that their sizing is optimal from the point of view of the level of service provided at the lowest cost, and without any assurance of the robustness of the supply chain.
Other techniques have been in vogue for some years, for example multi-level methods to optimise the allocation of stocks between different distribution levels. These are interesting but only cover part of the problem.
It is to understand this subject globally that the data scientists of Dillygence can help you optimize your stocks from a modeling of your flows, based on an experience of nearly thirty years of modeling and knowledge of the supply chain.
In conclusion, methods exist to optimise stock sizing, but they are still not well known and not used enough. Yet they would allow for more judicious and relevant stock sizing, enabling the stock to play its role, that of optimising production capacity and ensuring optimal service to customers.