Supply Chain View

I was interested to find an article in this month’s Logistics & Transport Focus headed “No more lean times: why inventory is not waste and warehouses add value”. The author, Steve Sordy, has chosen a title that is a kind of teasing of the more dogmatic of lean devotees – British culture has little patience for zealots of any stamp, and prefers to deal with them with dry irony.

Sordy makes some good points about value-adding activities that can occur in warehouses, such as late customisation, reformatting and returns processing. You could argue he muddies the waters a little by considering activities that could equally happen in stockless distribution centres, but overall it was a good summary.

Where I part company with the article is its assumptions – widely held misconceptions – about the role of inventory in lean supply chains.

Muda: the seven wastes

Inventory is indeed, as Sordy states, one of the seven forms of muda or waste identified by Toyota forty-odd years ago. Since Taichi Ohno and colleagues made that formulation, western popularisers such as Schonberger in the USA and Dan Jones in the UK have promulgated the idea that lean enterprises should be aiming for zero waste – and that means zero inventory. (Zero Inventories was for example the title of an early book on lean/JIT by Robert Hall in 1983.)

The New Romantics

But there is a difference between practical lean methods and “Romantic Lean”, that form of proselytising and sloganeering that is designed to win hearts and minds from the boardroom to the shop floor. Lean methods seem very counter-intuitive to people at all levels the first time they encounter them – reducing batch sizes for example, when there are long set-up times, or allowing machines and workers to idle if the downstream processes are blocked. So there is a need for this kind of advocacy. The danger is, of course, that people get confused by oversimplification.

Inventory in Pull systems

If there is a defining characteristic of lean production it is Pull. In manufacturing, pull means that production at any point is controlled by the status of the process downstream. The first and still most well-known way of doing this is Kanban.

In Kanban each process centre has a stock of material waiting to be worked on (a WIP queue). Each time a material is take from the queue, a Kanban card is sent to the upstream process authorising production. No process is allowed to start work on a job without a card. The number of cards are limited.

As the preceding description makes clear, holding inventory is actually an integral part of the lean pull process. Far from inventory being zero, it has become instituted in the system. In fact one of the ways of understanding pull is to think of it as a “make-to-stock” system, whereas push (e.g. traditional MRP) is a “make-to-order” system.

Furthermore, Little’s Law (Throughput = WIP / Cycle Time) shows that zero inventory really means zero throughput.

This means that lean manufacturing’s pull system leads to higher stock holding than push processes? No. In fact it leads to lower WIP and shorter cycle times, for reasons that are too complicated to explain here.

Don’t mention the “T” word

Sordy concludes that “only unnecessary inventory… is waste” and goes on to describe inventory holding decisions as “principally a series of trade-offs”. This is where the debate gets decidedly sticky.

It is true that it is important to understand the trade-offs, and it is true that only unnecessary inventory is waste. The problem is, what assumptions should we make when evaluating the trade-offs and what do we mean by unnecessary?

Sordy gives an example that inventory might be held because responsive manufacture “is an unaffordable option for many manufacturing processes that have significant changeover times”. The error in reasoning here is not the trade-off between stock and productivity, but that changeover times should be regarded as fixed. The lean view of batch inventory as waste leads us to reduce changeover times to change the terms of the trade-off. It was exactly this type of thinking that led Schonberger to ban the word “trade-off” (with unsuccessful results it must be said).

Behind the slogans of lean

Lean is not just about pull or Kanban. Neither is it about the seven wastes. As originally presented by Ohno, Shingo and Monden, it was in fact less of an overarching system than a general approach (an holistic view of the manufacturing process) together with a set of solutions to specific problems that had presented themselves over the years at Toyota. It was left to the consultancy and Business School synthesizers (the lean gurus) to give this the name “lean” and lend it some understandable shape (for example Womack and Jones’s five “Lean Principles”).

If you think about problems in a lean way (“lean thinking”?), it’s possible to see the trade-offs very differently. Little’s Law may mean we can’t attain zero inventory, but it also tells us that reducing cycle time and inventory together will maintain throughput – and reducing cycle time and inventory are both desirable outcomes, because they provide a more responsive service and lower costs.