Think in terms of value networks, not supply chains
FEATURE – This article explores the concept of “value network”, emphasizing the importance of looking at lead-times and value creation in a more holistic way.
Words: Michael Ballé, lean author, executive coach and co-founder of Institut Lean France
The first time I watched a Toyota sensei do a gemba walk in a plastic injection plant, he started with quality issues. The plant manager had put in place red bins to ensure his team checked non-quality parts coming out of the presses daily. He felt this was an extraordinary effort and complained corporate did not let him have the staff he needed to handle so many quality problems at once.
Through the interpreter, the sensei conveyed the idea that looking at quality once a day was simply ridiculous – no wonder the plant had so many quality problems. They needed to stop and investigate every defect, if they ever wanted to make money. Pointing at the hourly board that the plant manager had dutifully installed on the assembly cell, the sensei said: “This is the money board. Until every part is correct first time through, how can you expect to be profitable?”
You can imagine the rest of the conversation. “You don’t understand, we’re different.” “We don’t have the resources.” “What you’re asking is impossible.” And so on and so forth. The plant was mostly plastic injection and simple assemblies to make car components. At one point, the quality manager said, “Ok, we don’t know how to solve our quality problems. There, I’ve said it. Show us how!”
This gave the sensei pause. After thinking about it for a while he picked up a bad part from the red bin at the assembly station. “What is wrong with this part?” the interpreter asked. The quality manager looked at the component and pointed at one of the plastic parts and showed a surface defect on it. “Here – this won’t pass the customer’s inspection.”
“Where is this part made?” the sensei asked. The quality manager led the group to the injection press producing the parts. Sure enough, the red bin already had bad parts in it. The sensei looked at the press for a while, and then asked, “Where is the material kept?”
This took everybody by surprise. After some discussion, the group moved on to storage where large cardboard containers of plastic granules were stored, granules that would then be fed into the machine. “Have you checked the material?” asked the sensei. “Is it homogeneous throughout the load? Or do you have differences from one batch to the other?”
No one had thought of that before. A big quality fight in the plant was that press supervisors kept fiddling with the settings, adding instability to the process. No one had thought that maybe they had a reason for changing the settings: the fact that the material itself changed.
Value stream mapping is a great tool to get to know a plant. As you follow one product back through the process, you get a quick understanding of how it has been broken down into technical steps and develop the ability to distinguish value-adding time from stagnation time (all the time the semi-finished parts wait somewhere). The ratio of value-added time to total lead-time is usually stunningly low.
But this mapping also frames the delivery problem in a very particular way – focusing on one part, one client, one supplier. Yes, reducing batches at each station in order to reduce lead-time is essential. But that’s not the full story. In truth, rather than a value stream, we’re part of a value network.
A value network means that lead-time is not just caused by internal batching, but also by upstream demand scheduling: what demand do we expect from which customer? Which customer are we going to serve first? How are we going to schedule production? Downstream quality and delivery issues also have a big role to play: notice for instance how one tier-3 supplier supplies material for most of tier-2 – it might be a small, cheap component – but a shortage can still stop the entire chain, as we’ve recently seen with the global superconductor chip shortage.
With the pandemic, automakers slashed their orders as demand for chips in laptops and data centers increased. As the car and truck market rebounded, shortages started to appear with many components (our supply chains are not known for their reactivity). Note, also, that Toyota has been so far largely spared from the shortage, due to its better handling of its value network.
Indeed, just-in-time is not just about reducing stocks – it’s about carefully managing the steadiness of supply at each level in order to face head on business interruption risk. Frequent pull of components means frequent discussions with suppliers and, therefore, early discovery of problems – particularly when “bad news first” is the foremost request throughout the chain.
Granted, value stream maps are already hard enough to do and value networks simply won’t fit on the page – or in the mind. Nevertheless, we should not let the simplicity of the value stream image lure us into a false understanding – a misconception – of where and how lead-time is really created. Certainly, batch size is a central culprit, but demand scheduling and supplier integration are equally important to understanding, stabilizing and reducing overall lead-times.
Next time you’re in a plant, ask them how much of their value is purchased and go and have a look at their procurement process – you’ll see how the real point of pull in just in time is to establish a capillarity of value, capturing it drop by drop and making sure it keeps dripping, rather than brutally extracting it in large quantities – only to discover shortages and quality issues later in the process. Ultimately, value is produced by a full value network, not just by a sum of streams.