Dear Gemba Coach,
This is 2019. Isn’t there a better way to manage inventory than just-in-time by now?
Probably. Just-in-time is not an inventory management system. It’s a relationship management system that uses inventory as a wedge. Remember the old Star Trek movies (funny to think of science fiction in the past tense). They used replicators: push a button and your cup of tea materializes as you watch.Just-in-time was not an inventory control mechanism. It was a relationship building process, so each of the participants could react better to unexpected events and find together countermeasures to still maintain performance.
But still there is a lead-time. It’s almost instantaneous, but not quite (to be fair, it would ruin the lovely visual effect). There is always a lead-time. This lead-time is made of 1/ value added and 2/stagnation. Seen from the component’s perspective, every process is like the army: boredom punctuated by moments of rapid action.
Inventory shouldn’t be managed. It should be eliminated altogether. But of course, there’s a catch. Think of this column. If I have absolutely zero inventory, and start writing when my editor asks for it, he will have to wait during the time it takes me to write something – I don’t yet know how to write as fast as the replicator.
However, if I have one piece complete in time, I can send by return e-mail. Shorter lead-time.
A company that sells cars from a website learned just-in-time, improved its on-time-delivery from 50% overall to 80% and reduced its waiting time for customers (number one complaint) from two weeks to a week, which was much better than its competitors. They didn’t manage the inventory – they pulled the value.
Think: Relationship Building; Not Control
First, the sales points visualized their parking lots to see what car was going where (or waiting there). Then they learned to communicate more precisely with central logistics at headquarters about what cars to deliver the next day – and only the next day. Logistics then did the same with the depots to make sure the right cars were on the right truck. They also started giving trucks precise appointments and phoned them regularly to see how they were doing.
Some sales points played along, some didn’t. Some truckers played along, some didn’t. In the early days the difference in performance was stark, so eventually more people joined in. Just-in-time was not an inventory control mechanism. It was a relationship building process, so each of the participants could react better to unexpected events and find together countermeasures to still maintain performance.
Stagnation is the symptom of waste. I get a kick out of watching sensei walk a plant and talk to parts. Hello? How long have you been waiting here? Do you know where you’re supposed to go next? No one told you? Or when? A part sitting there has been manufactured or purchased – the cost has been spent. Then it has been moved – more cost. Then it’s using space and light – more cost. So, it’s doing nothing is costing you. That’s a waste. Muda.
Reducing lead-time is the way to reveal this muda and start eliminating unnecessary costs. The shortest way in, the wedge, is to reduce inventory without ever running empty and missing deliveries for customers. This should be easy to automate. Get an AI on the problem, they’ll solve it. Except that:
- Customers are fickle; they don’t always purchase the same quantities at the same time.
- Production schedules change so products or service jobs are not always done in the same quantity at the same time.
- Procurement is unpredictable, you never know exactly when you’re going to receive what – mostly because of the two previous reasons.
Computers can solve any predictable problems. But they’re really bad at dealing with unpredictable circumstances – which is precisely when humans shine. In a material process, there are two ways to reduce inventory: 1/ smaller batches, 2/more frequent deliveries.
Learning to reduce batches will obviously reduce inventory (well, not unless you deliver more frequently), but that’s not the point. First, it will get you to improve your quality – with the aim of producing the first part of the new batch right the first time without quality control (which I’ve seen in Japan on stamped parts, quite an achievement). Second, it will improve your flexibility and let you adapt to customer demand changes, and so make better use of your equipment, raising your OEE.
Supply Chain Capillaries
Learning to delivery more frequently will teach you rigor in logistics and flexibility in dealing with events deeper into the supply chain, which means frequent communication with suppliers and … relationship building. Quality issues can now be communicated faster and more often (when a faulty part is discovered, you can’t simply pick another one from inventory – there is no inventory. The next delivery comes with the next truck – so you call the supplier right away for them to check what they do) – which leads to closer engineering cooperation.
Just-in-time is in fact a way of drawing value in through capillarity, throughout the supply chain, like water flowing through a thread without being pushed through by pressure as water in a hose. By drawing work continuously from the supply chain, you not only improve delivery performance dramatically as well as reduce the total cost base. You also improve the relationships between all the participants, which leads to more innovative solutions and true innovations.
Truth is, changing from push to pull is not that hard. The benefits are incredible, as Toyota has been demonstrating time over time for half a century. So why do so few people go for it? They simply don’t understand it. The biggest obstacle to truly seeing just-in-time is thinking that it’s an inventory management thing. Inventory is the tool, not the goal.
Even in the lean world, I’m astonished at how so few people pull their own work and reduce just-in-time to some supply chain thing. Just-in-time thinking (getting all functions to cooperate by pulling work) and kanban (the tool to share the information) are the entry points to lean – practice them for yourself and all you know about lean will change.