A recent Wall Street Journal article set the lean world abuzz by claiming that automakers are retreating from 50 years of just-in-time manufacturing. I have heard something like this many times in the last 35 years, like every time there is a crisis that threatens Toyota’s supply chain. LEI has repeatedly challenged the widespread media mistakes in targeting JIT in the past, and this newest article even suggests that this ostensibly nonsensical practice will soon fade away. I’d like to re-examine and restate why just-in-time is so powerful and will remain ever more so moving forward.
While the pandemic has seen far too many shortages of everything from paper towels to testing availability and the vaccine itself, JIT is simply getting too much credit, and blame, for inventory shortages. Having studied Toyota (which is often viewed as the original model of JIT) for over 35 years, I believe that journalists who are now decrying the problems of JIT fail to understand what it is—or at least how it was developed and is viewed by Toyota.
The real role of inventory is to function as a buffer to enable flow of value to the customer.
One common misunderstanding among the many articles attacking JIT is the assumption that it functions primarily as a policy of controlling inventory. Hold as little inventory as possible to free up cash flow, this thinking goes, and if the inventory comes from a supplier, demand JIT deliveries of small batches frequently, and make the supplier hold the inventory and pay dearly if they shut down the customer’s production line.
This limited and transactional interpretation misses one fundamental point: Toyota does not view JIT as a standalone policy or program, but rather as an integral part of the Toyota Production System. Toyota views its core suppliers as partners. As Taichi Ohno put it:
“Achievement of business performance by the parent company through bullying suppliers is totally alien to the spirit of the Toyota Production System.”
The real role of inventory is to function as a buffer to enable flow of value to the customer. Any manufacturer receives direct shipments of materials to use in production. The orders for these materials were traditionally scheduled based on forecasted demand, which often led to a mismatch between what is ordered and what is needed. Toyota chose to schedule these inbound deliveries through the use of a pull signal from the production line, known as kanban. The process that needs these materials looks at the actual condition and, based on rules, orders inventory to be replenished when needed. A simple binary signal—I am ready for more or I am not—puts responsibility for ordering in the hands of team members who know their situation.
Pull systems operate in a similar way to scheduling. At minimum, enough materials must be ordered to cover the replenishment time. For example, if you get a shipment once per week you must have at least one week of parts on hand to cover this lead time. If the parts are shipped from overseas and it takes four weeks, you need a minimum of four weeks of inventory.
When things go wrong in the production and shipment of goods, we need some extra inventory to cover that eventuality.
This inventory serves as a buffer against the time it takes to replenish and will work as long as everything goes exactly according to plan. But it never does. There will always be variability—a factor that we have seen frequently in recent times. And when things go wrong in the production and shipment of goods, we need some extra inventory to cover that eventuality. That emergency supply is called safety stock. If the maximum time you have experienced in your weekly deliveries can sometimes be 14 days from the time you place the order you might want to hold 14 days of inventory—7 days for average replenishment plus 7 days of safety stock.
JIT done right also accounts for the variability caused by customer demand. If customer demand suddenly doubles from the average rate, you would need to have the original 7 days of inventory, plus 7 days for the maximum demand, plus the 7 days of safety stock—a total of 21 days.
On top of that, you will always face the possibility that your equipment will break down or you will be shut down by material shortages from your supplier. You must add enough to your buffer to cover these contingencies as well.
Hopefully the picture is getting clearer: just-in-time done right is still far from a zero-inventory system. Moreover, pull systems with kanban are not immune to lead times and variations on the times. The real takeaway from a more detailed look at this should reveal that variability is the enemy of JIT.
The Roots of JIT
When Toyota Motor Company was first formed in 1933, Kiichiro Toyota was starting from scratch. As in many startups they did not have enough cash to capitalize the company. Nor did they have suppliers; they did not even have customers. Once they set up production and got some customers they still did not have enough cash to run the company. The pressure was to build the product fast and get paid fast so Toyota could pay employees and suppliers. Toyota could not afford to have money tied up in inventory and needed lead times as short as possible making JIT a necessity. As Taiichi Ohno later said: “All we are doing is looking at the time line, from the moment the customer gives us an order to the point when we collect the cash. And we are reducing that time line.”
Kiichiro Toyoda developed the idea of JIT and, being very detailed oriented, developed a four-inch binder that described in meticulous detail how the system should operate. It was the foundation for the kanban system. Kanban was not nirvana, but rather was a compromise. In the ideal situation products would flow down an assembly line or through a cell one piece at a time without inventory. But this was not possible for batch processes like stamping, or for parts coming from suppliers some distance away. So when inventory was necessary it was controlled with kanban.
The barriers to smooth flow were various forms of variation.
This was all well and good, but kanban did not solve the root cause of the problem. As we saw with safety stock, the barriers to smooth flow were various forms of variation. This could be variation in customer demand, variation in the speed of the production line, variation in manual tasks performed by people, variation in the functioning of equipment, variation in traffic for deliveries, variation in the supplier’s internal processes, and more. Of course no company, including Toyota, can control everything, so variation was inevitable. But one characteristic of the Toyoda family going back to Sakichi Toyoda was perseverance. They would work, experiment, and struggle to control everything they could to make the value stream stable.
The full production system needed to make JIT work and flow value to the customer was ultimately represented by the TPS house, a system. While JIT was one pillar, the corresponding pillar was jidoka. Jikoda referred to the process of quickly identifying any abnormality and stopping to fix it. The first step was containment to get production going again, and the next step was root cause analysis to try and eliminate the problem. This led to the famous andon system where a machine or worker identifies problems in real time and stops the process. The purpose was not to keep shutting down production, but rather to identify and correct deviations from standard conditions as they occur. And who could do this with the thousands of processes involved? The answer was team members and their team leaders and group leaders—those responsible for the process. Thus everyone had to be developed as problem solvers.
Even stopping to fix problems was not enough. If the Toyota assembly plant was on a roller coaster ride, chasing variable customer demand there would be too much variation to control and the well-known “bull whip” effect would wreak even more havoc with suppliers and suppliers to suppliers. The further back you go the more disruptive are changes you will see in Toyota’s takt or rate of customer demand. The only solution would be to hold more and more inventory as you move backwards in the supply chain. So to show respect for suppliers and enable a smooth flow of work Toyota invested in heijunka, or production smoothing.
To show respect for suppliers and enable a smooth flow of work Toyota invested in heijunka, or production smoothing.
Heijunka was depicted as the foundation of TPS and means Toyota does not build things following the sequence of customer demand but takes demand over a period of time and smooths it out. This allows for a stable takt and creates a target for the entire production process—build to the average of customer demand. Of course Toyota could not smooth all variation in customer demand so they developed a flexible, adaptive system. This included building multiple products on the same assembly line so that while demand for an individual model could go up and down the average demand across vehicle models would be more stable. It also included having a flexible work force to add or subtract work hours depending on demand. For example, by running two shifts instead of having round-the-clock work, there was time between shifts to add or subtract overtime.
In this way the full Toyota Production System evolved and Toyota’s learning evolved. For example, Ohno later learned that there were benefits to JIT beyond having less inventory and getting paid faster. In his famous analogy of lowering the water level to reveal the rocks he discovered that excessive inventory removes pressure to improve processes. After all, if your supplier or preceding process shuts down, you can continue working as long as you have inventory. If on the other hand a supply stoppage threatens to shut down assembly imminently, you will feel intense pressure to keep the line running, and to solve problems that occur at the root cause. This is how a system of connected processes with little inventory encourages and even demands continuous improvement. In fact JIT was a way to surface problems and support continuous improvement which stabilized processes and supported JIT.
My recent second edition of The Toyota Way defines JIT as: “A system of continuous flow that brings all materials and information in small lots to the point of use as they are needed—neither too early nor too late. This avoids waste including overproduction and creates a more efficient flow that quickly surfaces abnormalities so people can improve quality, cost, on-time delivery, and responsiveness to changes in customer demand.”
JIT is Not the Answer to Every Crisis, or the Cause
Even with decades of practice with and learning about the power of JIT, Toyota is still not immune from serious parts shortages today. Just as you can’t squeeze water from a rock, Toyota cannot make cars out of non-existent parts. And Toyota cannot meet demand when hit with sudden surges in demand that Toyota did not anticipate.
The challenges faced by a run on supplies—such as toilet paper in the initial stages of COVID—creates problems for everyone. If critical parts stop flowing to Toyota plants because of a natural disaster or a breakdown in a supplier plant or the transportation system, Toyota will indeed run out of parts soon, and faster still if that is a JIT supplier and there is little inventory in the system. There is an easy answer to this—hold six months of inventory of everything and become a warehousing business instead of a manufacturer.
Toyota consciously chooses NOT to pursue that policy. Doing so would wipe out all the benefits of JIT that I’ve described. Instead, consider this countermeasure: hold the amount of inventory necessary to cover production and delivery time in normal circumstances, while carrying sufficient safety and buffer stock to cover variability. In this case the inventory policy will vary with each item. Toyota accounts for this by having a “plan for every part.” That’s why highly reliable suppliers whose plants are located close to the Toyota plants require minimal quantities of inventory, while overseas suppliers, or suppliers who do not have their processes under control, call for greater amounts. Even more inventory is needed if there is a forecasted disaster, like a tsunami on the horizon. And Toyota hedges its bets by having two or three sources for each part in each region of the world and seeks to have the shipping points geographically dispersed within a region.
Indeed, even the WSJ Journal article oddly conceded that Toyota itself (the company most synonymous with JIT) has adapted its use to respond successfully to extreme situations. The later approach of multiple, dispersed suppliers has been a lesson Toyota has learned many times. For example, the 2011 Fukushima earthquake that led to a nuclear power plant meltdown shut down suppliers in the central region of Japan. One of these shuttered suppliers was a computer chip supplier which served multiple customers. This was Toyota’s sole source of these parts so when that one plant went down and shut down production for months globally Toyota had no chips. This was not a failure of JIT and in fact this was not a JIT supplier. It was a failure of Toyota following its own supplier strategy. The lessons: have multiple, geographically dispersed production sources, and when a source is at risk, hold inventory. They did both these things, including asking suppliers to hold weeks of chip inventory, and as a result have benefited greatly in the 2021 chip shortage which has shut down many competitors.
Toyota has adapted its use of JIT to respond successfully to extreme situations.
Even if Toyota had failed to improve its JIT approach and faced months-long shutdowns in the face of periodic shortages, it would have never abandoned JIT in the context of TPS. Striving for one-piece flow without waste is core to the business philosophy of Toyota. This philosophy does not change with a crisis. Instead, Toyota deals with a crisis by doing deep reflection on what is to be learned and adjusting.
When there is a crisis Toyota also benefits from its close partnership with its highly capable first-tier suppliers who come together to resolve the problems as quickly as possible. This was evident in the famous 1997 Kobe earthquake when the sole source of a p-valve needed for all brake systems shut down its operations. Only 2-3 days of stock was available and a global shutdown seemed imminent. Yet disaster was averted as over 200 suppliers self-organized to develop capacity to build the p-valves and supply Toyota within days.
What Can Companies Outside the Toyota Group Do With JIT?
Unfortunately one thing that legitimately drove the WSJ article is the large number of companies that have adopted JIT deliveries as a policy to reduce inventory—without using any of the supporting systems of Toyota. Simply implementing this in isolation merely as a tool is likely to cause many problems. Remember, one benefit of JIT is that problems are quickly surfaced so they can be solved. Surfacing problems is not desirable for a company that does not have highly developed people and suppliers throughout the value chain who are continuously improving.
One benefit of JIT is that problems are quickly surfaced so they can be solved.
Companies that lack mature lean systems and have not sufficiently developed their people can reap limited benefits of JIT, and the use of a tool like kanban can help with inventory control. But the primary enemy will always be the hazards of managing variation. And, the more variation you have, the more inventory you will need as a buffer.
Most importantly, Toyota learned to stabilize processes and concurrently reduce inventory for those cases where there is a reliable stream of parts and materials. They have shown that working with key suppliers who become partners will enable more reliability and less inventory. Working to stabilize one’s own processes in the direction of continuous flow to takt will also reduce the amount of inventory needed. And, working relentlessly on the difficult problem of leveling the production schedule, or heijunka, will have enormous benefits, particularly as one works back through the tiers of suppliers. This is not to say that Toyota’s solutions are to be copied. Each company is in a unique situation and needs to strive to discover its own approaches to move in the direction of a pure stream of value to deliver what customers want, in the amount they want, when they want it.
This all takes time and a great deal of effort. Toyota has done this work and enjoys the benefits. The company understands the need to hold inventory when they anticipate a looming crisis or when there are no reliable sources of materials. Yet these are episodic instances when deviations from broader practice are called for. It’s more important to understand the purpose of JIT and where it fits into a deeper system. Companies and journalists should not pretend JIT is possible when it is not, or that it is impossible when it is. And they should focus instead on striving in the direction of a lean supply system to enjoy the powerful benefits it delivers.