A Single Door Check Stops Japanese Cars: Why 10 Years of Preparation Failed in the Re-disaster in Kumamoto

On the evening of July 28, 2026, a major earthquake struck Kumamoto once again. It was a magnitude 7.1 quake with a maximum seismic intensity of 7. In this region, where memories of 10 years ago are still vivid, the factory of automotive parts manufacturer Aisin Kyushu was hit by disaster for the third time. When this factory stops, the assembly plants of Toyota Motor, Nissan Motor, and Daihatsu Motor stop. And indeed, they did stop.

The problem is that it is not as if this company had not prepared for anything. After the disaster 10 years ago, seismic equipment was installed on ceiling-suspended piping to suppress shaking, and stoppers were attached to mold shelves. Five days' worth of inventory was secured, and a system was established to produce the same products at multiple factories. Even so, the third major earthquake shattered these "10-year preparations."

In this article, after organizing the facts of what happened in Kumamoto, I will delve into why the 10 years of preparation were insufficient, moving from the micro level of parts to the production base of Kyushu, and finally to the macro level of Japan's supply chain design. This is a story about disaster prevention, but it is also a point of discussion for manufacturing management, risk management, and investors looking at Japanese stocks.

1. July 28, 2026: The Earth Shakes Again in Kumamoto

The earthquake occurred around 16:27 on July 28, 2026. The Japan Meteorological Agency announced the epicenter was in the Kumamoto region of Kumamoto Prefecture, with a magnitude of 7.1. The depth of the epicenter was reported as approximately 10 kilometers in the initial bulletin, and later reported as approximately 16 kilometers.

The maximum seismic intensity of 7 was observed in Uki City and Hikawa Town, Kumamoto Prefecture. A seismic intensity of 6-upper was recorded in Yatsushiro City and Uto City, and 5-upper was recorded in Kikuyo Town. The government issued a tsunami advisory for the Ariake Sea and Yatsushiro Sea, predicting a tsunami of up to about 1 meter.

As of the report on the day after the earthquake, 13 deaths had been confirmed. An event widely reported as symbolizing the damage was an explosion at the Aeon Mall Kumamoto in Kashima Town. It is reported that the explosion occurred about an hour after the evacuation of customers and employees had proceeded following the earthquake. The fire department received a report that "white smoke was rising and there was an explosion sound," and the possibility of a gas leak was being investigated at the scene. At the time of the report, 3 deaths were confirmed in this explosion, 1 person was in cardiopulmonary arrest, 4 were transported to the hospital, and 3 employees were reported missing. It is reported that parts of the second floor, exterior walls, and roof of the building collapsed.

Whether the cause of the explosion was damage to piping due to the earthquake or ignition of leaked gas was not confirmed at that time. This is a part where one should avoid jumping to conclusions. However, the fact that human lives were lost in a way different from the collapse caused by the direct shaking of the earthquake shows once again that the damage from a disaster does not end the moment the shaking stops.

Regarding damage to industrial facilities, the collapse of a chimney at the Yatsushiro factory of Nippon Paper Industries was reported. Power outages occurred in approximately 47,210 households in the prefecture. The Shinkansen temporarily suspended operations, and some roads were closed to traffic.

And the place that became the most symbolic of the industrial impact of this earthquake was a single parts factory in the southern part of Kumamoto City.

2. The Third Disaster: What Happened in Southern Kumamoto City

The Aisin Kyushu factory in the Jonan district of Minami-ku, Kumamoto City, has now been hit by a disaster for the third time.

The site area is 220,000 square meters. Three buildings are largely connected. What is produced here is the "door check," a core component that controls the opening and closing of car doors. At one time, this factory was responsible for the entire volume of Toyota's domestically produced vehicles.

Regarding this disaster, an official from Aisin Kyushu confided, "I think it's worse than last time. We haven't even been able to test if the production equipment works in the first place." It was reported that all employees on duty were safe, and while no major damage to the buildings themselves was confirmed, detailed damage inside the buildings was still being checked. There is no prospect for resuming operations.

What should be noted here is the content of the damage. According to reports, a part of the ceiling-suspended piping ruptured. Although stoppers were attached to the mold shelves, "a mold for one product number" fell from the shelf. And the device that cools the production equipment was in a state where it could not be operated, leaving the factory at a stage where they "had not yet been able to test" whether the production equipment would actually function.

In other words, the building remains. The people are safe. Yet, the factory does not move. This asymmetry is the essence of this event. Even if you only increase the seismic resistance of the building, the factory will not move.

3. The True Nature of the Part Called a Door Check

Many people have probably heard the word "door check" for the first time. It is not listed in catalogs or specification sheets. Even so, if just one of these parts is missing, the finished car assembly line will stop.

A door check is a component that holds the door at a certain angle and provides appropriate resistance and moderation during opening and closing. The function of keeping the door stopped in a half-open position, and the "click" feeling when it stops opening, are handled by this part. The reason a car door parked on a slope does not open forcefully and hit the car next to it is because this part is working.

This part has several troublesome characteristics.

First, it is directly linked to safety and quality. If a door were to open fully unintentionally, passengers could fall out or adjacent vehicles could be damaged. Therefore, it is a part for which it is difficult to find a workaround like 'assembling it without it for now and attaching it later'.

Second, the shape differs for each vehicle model. Because they are designed according to door weight, hinge angle, and body rigidity, they are not general-purpose products. This is the meaning behind the reports that what fell from the shelf was 'a mold for a single part number.' Even if only one mold fell, production of the vehicle model corresponding to that part number will stop.

Third, the unit price is low. As a part, it is inexpensive, which is precisely why it is structurally easy to consolidate production bases from the perspective of profit margins. If you disperse the production of cheap parts, profitability collapses due to duplicate investment in molds and logistics costs. The accumulation of rational management decisions results in 'single-point concentration'.

The combination of these three characteristics creates a structure where 'cheap and inconspicuous parts become a bottleneck for the entire industry.' What the Japanese automotive industry learned 10 years ago was precisely this structure.

4. Memories of 10 years ago: The day 90% of Toyota's domestic production stopped

In April 2016, this factory suffered catastrophic damage from two major earthquakes.

The description of the damage is specific. Almost all of the ceiling-suspended piping fell. Shelves along the walls tilted, and molds broke through the walls and flew outside. A Nikkei reporter who covered the scene at the time saw workers clearing away the debris of the walls, trucks hauling away broken equipment, and three cranes looking down on the factory. Voices were heard from employees saying, 'The equipment has fallen over and it's a mess.'

And the shutdown of this one factory forced the almost total suspension of production for Japan's largest manufacturing company.

The two factories, Aisin Kyushu and Aisin Kyushu Casting, entered a shutdown following the earthquake that occurred on the night of April 14. As of April 16, they were proceeding with safety checks of the buildings, but it was taking time to confirm damage to production equipment, and there was no prospect for restarting operations.

On April 17, Toyota announced that it would gradually suspend domestic finished vehicle assembly lines from April 18 to 23 due to the stagnation of parts supply. The Toyota Motor Kyushu Miyata Plant stopped on April 18, and from the 19th, main plants in Aichi Prefecture such as Takaoka, Tsutsumi, Tahara, and Motomachi, as well as Toyota Auto Body plants and Toyota Motor East Japan's Higashi-Fuji plant, stopped sequentially. From the 20th to the 23rd, all of Toyota's mass production lines were suspended, with the suspension target reaching 26 out of 30 domestic lines. The expression that about 90% of domestic finished vehicle production lines stopped refers to this figure.

The impact on the production side was also significant. Toyota initially expected a total production reduction of about 50,000 units, and other reports stated that the suspension resulted in a total production delay of 80,000 units. The domestic finished vehicle lines fully resumed on May 6. That is a gap of more than three weeks from the disaster. As of the end of April, domestic production had returned to a scale of 10,000 units per day, which is about 80% of the pre-earthquake level, and returned to the pre-earthquake level of about 13,000 units per day after May 23.

The measures taken by Aisin at this time are also recorded. According to the company's announcement on April 28, 2016, for Aisin Kyushu, production equipment and molds were removed from the factory, and alternative production began in stages from April 23 at Aisin Seiki and group companies in Aichi Prefecture, as well as partner companies in the Kyushu region. It was planned to start alternative production for 'almost all items' by May 2. Aisin Kyushu Casting resumed production in stages from April 27 and was scheduled to resume mass production in all processes on April 29.

Without waiting for the restoration of the damaged factory itself, taking out the molds and making them elsewhere. This decision was the core of the lessons learned 10 years ago.

5. What has changed in 10 years? The content of the measures taken

When a Nikkei reporter visited this factory again in April 2026, it was written that the large holes in the walls had been filled and the memories of the earthquake had become distant. However, measures had certainly been implemented.

The measures can be understood by dividing them into three major layers.

The first layer is the physical securing of equipment. Seismic devices were installed on the ceiling-suspended piping to suppress shaking. 'Stoppers' were attached to the mold shelves, and measures were taken to prevent the molds from sliding off. This means that measures were taken one by one for the things that fell and flew out 10 years ago. At the entrance to the factory office, a collection of messages of support received during the disaster 10 years ago is still displayed, and the words 'Don't give up, Kumamoto' remain. It was also a device to keep the memory from fading.

The second layer is the thickness of inventory. According to those involved, based on the lessons from the previous time, they had secured 5 days' worth of factory inventory, which was almost non-existent before. In the Japanese automotive industry, which advocates just-in-time, the decision to hold 5 days' worth of inventory is itself a major policy shift. It means that the workplace, which had made a virtue of zero inventory, has steered in the direction of intentionally accumulating inventory.

The third layer is the multi-tracking of production bases. A system was established to produce the same product at multiple factories. Even if one base stops, it can be replaced by another base. 10 years ago, it took 9 days from carrying out the molds to starting up alternative production, but if you have already been running them at multiple bases, the switching time can be compressed.

Furthermore, as an industry as a whole, moves to multi-track parts logistics networks and efforts to visualize the supply chain back to the nth-tier subcontractors have also progressed. What Toyota worked on after the earthquake was building a mechanism to grasp which parts are made where, including second- and third-tier suppliers with which it does not trade directly. It is the idea of having a map in advance of the paths through which the disaster of one company spreads.

When you list them out like this, these measures fall into the category of being quite serious. Even so, the factory stopped.

6. It stopped anyway: The third tremor that broke through the countermeasures

It is reported that the third major earthquake seems to have broken through even these measures. Let's look at what happened by matching the countermeasures with the results.

Seismic equipment was installed on the suspended piping. Even so, part of the piping ruptured. Devices that suppress shaking reduce the amplitude of the vibration. However, they do not eliminate the internal pressure of the piping or the stress applied to the joints themselves. It cannot be denied that by being fixed with equipment, force may have actually concentrated on a specific point.

Stoppers were attached to the mold shelves. Even so, a mold for one specific part number fell off the shelf. Stoppers stop the displacement of molds within the range of assumed acceleration. They cannot handle shaking that exceeds assumptions or shaking in directions that were not anticipated. And, only one fell. Even if it is just one out of the total, production for that part number stops.

And another type of damage pointed out by reports is highly suggestive: the fact that the equipment used to cool production machinery could not be operated. In the molding process using molds, cooling determines quality and production speed. Cooling equipment is not the "production equipment" itself, but rather the utility equipment surrounding it. The priority for seismic countermeasures is inevitably higher for the main equipment. However, if the utilities do not work, the main equipment will not work either.

From this, we can read the following three points.

The first is the nature that countermeasures are taken against "what fell last time." Fixing the piping that fell last time and stopping the mold that flew out last time. This is rational learning, but it is essentially reactive. What will break next is not on the list of what broke last time.

The second is that even if the effect of the countermeasures turns a "catastrophe" into "partial damage," the result of a production stoppage remains the same. Ten years ago, holes were punched in the walls of the building. This time, no major damage to the building has been confirmed. The degree of damage is clearly lighter. Even so, the factory has stopped. The fact that those involved said, "I think it's worse than last time," should be seen as referring not to the scale of the damage, but to the uncertainty of not having been able to test whether the production equipment would actually run.

The third is that the bottleneck of factory operation is determined by the "weakest single point." Three buildings, a 220,000-square-meter site, and countless pieces of equipment. If a mold for one part number and one cooling device among them cannot be used, that part stops. Even if you protect 99% of the entire factory, the remaining 1% dictates production.

7. The map of ripple effects: The status of stoppages at Toyota, Nissan, and Daihatsu

Products made at this factory are supplied not only to major customer Toyota, but also to Nissan and Daihatsu Motor. Since the situation where supplied goods cannot be produced will continue after the 30th, it is expected that the assembly of finished vehicles by the three companies will also be affected.

Let's organize the specific stoppage status.

On July 29, Toyota decided to stop operations at three factories in Fukuoka Prefecture until the 31st. These are the Miyata, Kokura, and Kanda plants of Toyota Motor Kyushu. Although these three factories resumed operations once on the morning of the 29th, they followed a course of stopping again from the evening of the 29th until the 31st. The reason for the stoppage was explained as being due to considerations for safety confirmation, logistics, and the situation of suppliers. Operations from August 3rd onwards were left undecided. The Miyata plant is a production factory for the luxury car brand "Lexus," and cars made here are exported to China and other countries.

Nissan will stop some production lines at its subsidiaries, Nissan Motor Kyushu (Kanda Town, Fukuoka Prefecture) and Nissan Shatai Kyushu (same location), from the 30th to the 31st. Nissan produces its flagship SUV, the "Rogue," in Kyushu and transports it to North America. It is a flagship model for the North American market, and the significance of this stoppage is not small.

Daihatsu also announced that it would stop operations at its subsidiaries Daihatsu Kyushu's Oita Plant (Nakatsu City, Oita Prefecture) and Kurume Plant (Kurume City, Fukuoka Prefecture) until the 31st. According to reports, the Oita Plant resumed operations once after safety confirmation, but the Kurume Plant remained stopped, and the stoppage is set until the 31st.

What we want to look at here is that the reasons for the stoppages are not uniform. The stoppage immediately after the earthquake was due to employee safety confirmation and building inspections. This can be determined in a few hours to a day. The reason Toyota's three factories resumed once on the morning of the 29th was because this safety confirmation was completed.

However, they stopped again from the evening of the 29th. The reason this time is parts supply. A stoppage due to safety confirmation and a stoppage due to parts supply are completely different in nature. The former can be solved by one's own efforts. For the latter, there is no choice but to wait for the recovery of the affected business partners or to set up alternative supplies.

And a stoppage due to parts supply has the characteristic that the end is difficult to see. The expression "operations from August 3rd onwards are undecided" should not be read as simply meaning it hasn't been decided, but as meaning it cannot be decided at a stage where it has not yet been tested whether the production equipment at Aisin Kyushu will work.

Damage outside the automotive industry has also been reported. Renesas Electronics stopped its Kawashiri and Nishiki plants in Kumamoto Prefecture; at the Nishiki plant, ceiling panel falls and wall cracks were confirmed, and at the Kawashiri plant, wall cracks and water leaks were confirmed. Mitsubishi Chemical Group's Kumamoto plant has broken windows and other damage, and resumption is undecided. The Uki plant of automotive parts maker Astemo cannot foresee a resumption due to power outages and restricted access, and employee injuries have also been reported. Honda's Kumamoto factory also remains stopped.

8. Kyushu as a Production Base, Supporting Over 10% of Domestic Output

Kyushu is a core hub for automobile production. According to the Kyushu Bureau of Economy, Trade and Industry, automobile production in Kyushu is 1.24 million units in 2025, accounting for over 10% of domestic production.

I will confirm the significance of this figure from several perspectives.

Kyushu is home to four finished vehicle manufacturers, with an annual production capacity reaching approximately 1.54 million units. Its national share of four-wheeled vehicle production is considered to be about 15%, positioning it as one of the world's leading production bases. After declining from fiscal year 2019 due to the impact of parts supply shortages, Kyushu's four-wheeled vehicle production recovered to 1.22 million units in fiscal year 2022. The figure of 1.24 million units for 2025 is an extension of this recovery.

The ratio of 'over 10%' might not seem large as an absolute value. However, automobile production cannot be discussed solely in terms of unit ratios. There are three reasons for this.

First, the export ratio is high. Toyota Motor Kyushu has announced that it exports 90% of its produced vehicles overseas. Lexus vehicles made at the Miyata Plant are shipped to China and other regions, while the Nissan Rogue is transported to North America. The ratio of over 10% of domestic production carries a different weight from the perspective of exports. If production in Kyushu stops, the impact will extend to Japan's trade balance and even to macro indicators through exchange rates.

Second, vehicle models are difficult to substitute. Lexus is not a brand whose models can be easily substituted by production at other plants. Luxury car brands have strict process management and inspection standards, and transferring production takes time. The Rogue is also an SUV exclusively for North America, so the option of shifting production to another domestic plant is not realistic. It is important that the content of this '10%' consists of highly profitable strategic models rather than replaceable mass-produced cars.

Third, there is the agglomeration effect on employment and the regional economy. Around each finished vehicle plant, numerous primary to tertiary parts manufacturers are located. The automobile industry in Kyushu is a major pillar of regional manufacturing employment. Looking only at the manufacturing industry in Kumamoto Prefecture, there is a scale of 2,235 business establishments, 93,807 employees, 3.4863 trillion yen in manufactured goods shipments, and 1.2588 trillion yen in value-added. The suspension of a plant directly affects not only the daily production volume but also the household finances of the region.

9. Silicon Island Shaken Simultaneously: A Double Bottleneck

There is a point where this Kumamoto earthquake differs decisively from the one 10 years ago. It is the fact that in these 10 years, Kumamoto has transformed into a semiconductor hub.

The value of manufactured goods shipments for Kumamoto Prefecture's semiconductor-related industries (semiconductor manufacturing equipment, integrated circuits, semiconductor elements) reached 991.9 billion yen in 2023. As of June 2024, the number of employees is 16,892. This industry has expanded to account for 28.5% (shipment value) and 18.0% (number of employees) of Kumamoto's entire manufacturing industry. The number of location agreements between Kumamoto Prefecture and semiconductor-related companies has reached 70 in total since the announcement of TSMC's entry.

The number of employees at the TSMC Kumamoto Plant 1 (JASM), which began mass production in December 2024, was approximately 2,400 as of April 2025. According to Ministry of Finance data, the investment amount related to TSMC, Sony, Denso, and Toyota is shown as 207.2322 billion yen (as of December 2024). An industrial cluster that did not exist during the Kumamoto earthquake 10 years ago was exposed to the same shaking this time.

The responses of various semiconductor companies were reported between July 29 and 30.

JASM's (TSMC) Kumamoto plant evacuated employees, conducted safety checks on the buildings, and resumed operations in stages after confirming structural safety. Although the second plant under construction did not suffer direct damage from the earthquake, some construction work was temporarily suspended in light of the risk of aftershocks.

Sony Semiconductor Manufacturing's Kikuyo plant continued to suspend production from the 28th, and as of the 29th, the resumption date was undecided. As mentioned above, Renesas stopped its Kawashiri and Nishiki plants, but it was reported that there were no problems with air conditioning or power supply equipment, and the clean rooms were maintained. Mitsubishi Electric also stopped operations at its plants in Koshi City and Kikuchi City to proceed with inspections, and it was reported that Tokyo Electron Kyushu, TOPPAN, and Fujifilm-affiliated bases also began suspension and inspection procedures out of caution for aftershocks.

The view of experts and industry sources is that while no large-scale equipment damage has been confirmed at this time, there is a concern that if aftershocks continue, the stagnation of semiconductor production in Kyushu will spread and affect the global supply chain. Within the scope of reports obtained, economic loss estimates on a monetary basis have not yet been released.

What emerges here is the composition of a double bottleneck. Finished automobile production requires both mechanical parts (door checks) and semiconductors (automotive semiconductors, image sensors). The supply sources for both are in the same prefecture, within the same shaking.

Semiconductor plants have different risk characteristics than mechanical parts plants. If the clean room environment is disturbed, the time until production resumes is prolonged. Precision equipment such as exposure systems requires recalibration even with slight positional shifts. Conversely, if there is no structural damage and the clean room is maintained, it can be restored relatively quickly. The fact that the Renesas report specifically mentions that 'the clean room is maintained' is because this is the turning point that determines the speed of recovery.

The automobile industry experienced a semiconductor shortage phase from 2021 to 2022. The lesson from that time was to 'keep thick semiconductor inventories.' This time, it showed the possibility that mechanical parts and semiconductors could be threatened simultaneously, and by the same regional risk. In designing risk diversification, 'diversification by parts category' and 'geographical diversification' are different axes, and both must be satisfied simultaneously.

10. The Meaning of the Phrase 'The Impact Will Remain Within Kyushu'

Stakeholders also point out that "the impact will remain within the Kyushu region." This single statement is the most important part that shows the difference from 10 years ago.

Ten years ago, the disaster at one Kumamoto plant spread to Takaoka, Tsutsumi, Tahara, and Motomachi in Aichi Prefecture, and even to Higashi-Fuji in Shizuoka Prefecture. 26 out of 30 domestic lines stopped. This time, the ones that stopped were three Toyota plants in Fukuoka Prefecture, two Nissan plants in Fukuoka Prefecture, and two Daihatsu plants in Oita and Fukuoka Prefectures. The plants in Aichi Prefecture did not stop.

What created this difference are the three measures seen in the previous section: 5 days of inventory, production of the same products at multiple locations, and the diversification of logistics networks. The 10 years of preparation have certainly proven effective in the form of "reducing the scope of impact."

In other words, the accurate assessment is as follows: the 10-year preparation failed to prevent the plants from stopping. However, it succeeded in preventing the impact from spreading nationwide. If you look only at the former, "the measures were insufficient," but if you look at the latter, "the measures are working."

This distinction is critically important when considering risk management. A company's BCP (Business Continuity Plan) has two main goals. One is to protect the affected site itself and restore it quickly. The other is to stop the shutdown of the affected site from spreading to the whole. Let's call the former prevention and the latter containment.

Prevention, when pushed to the limit, becomes a battle against the forces of nature. Protecting equipment 100% against a magnitude 7 earthquake is not realistic, both as a cost issue and a physical issue. On the other hand, containment is a design issue. How much inventory to keep and where, how many locations to produce the same parts in, and how many logistics routes to have. This can be decided by management judgment.

What has truly made the Japanese automotive industry stronger over the past 10 years is not prevention, but containment. And it can be said that this was the correct order of investment. Rather than pouring infinite costs into prevention, investing in containment design yields higher investment efficiency.

However, this assessment comes with a condition: the condition that "restoration work can be completed while the inventory lasts."

11. Is 5 days of inventory a lot or a little?

Let's calmly evaluate the figure of 5 days of inventory.

First, from the principle of just-in-time, 5 days is a clear change in policy. The Toyota Production System has defined inventory as "waste." Inventory ties up working capital, creates storage costs, and delays the discovery of quality problems. The decision to hold 5 days of inventory means that they have explicitly decided to pay an insurance premium against the low-frequency, high-impact risk of an earthquake.

Next is the meaning of the length of 5 days. The disaster occurred on July 28, and it was stated that the situation where delivered goods could not be produced would continue after the 30th. Toyota's three plants stopped until the 31st, and the situation after August 3rd remains undecided. Calculating from the disaster, the 5 days of inventory will run out around August 2nd. This includes the weekend from July 31st to August 2nd.

When stakeholders say that "the weekend work is the climax," they are making statements with this timeline in mind. Can restoration work be completed while the inventory lasts? Can they find partners for alternative production as they did last time? These two questions are being asked.

Let's compare it with the track record from 10 years ago. In 2016, the disaster occurred on the night of April 14, and the gradual start of alternative production was April 23. It took 9 days. Since the start of alternative production for almost all items was set for May 2, that was 18 days after the disaster.

This time, the inventory is for 5 days. If the pace of launching alternative production from 10 years ago is reproduced as is, the inventory will not be enough. Conversely, the inventory level of 5 days should be designed on the premise that "the same products are already being produced at multiple locations in advance." Instead of moving molds and setting them up, they increase the production volume of lines that are already running. If that is the case, 5 days is a reasonable level.

The point derived from this is clear. The appropriate level of inventory is not determined in isolation. It is only determined in combination with the launch speed of alternative production. A combination of 5 days of inventory and an alternative line that starts in 3 days is safe, but a combination of 5 days of inventory and an alternative line that takes 9 days is dangerous.

This is one of the reasons why many companies' BCPs do not function. The target for inventory days and the Recovery Time Objective (RTO) are decided separately by different departments. Finance wants to reduce inventory from the perspective of working capital. Production engineering wants to shorten recovery time. However, if there is no mechanism to verify whether those two numbers are consistent, the inventory just becomes "5 days for some reason."

There is one more point that is easily overlooked. 5 days of inventory is "factory inventory." If this inventory is inside the affected factory, and access to the building is restricted due to the earthquake, the inventory cannot be retrieved even if it exists. The measure of holding inventory only makes sense on the condition that the inventory is "placed in a location that will not be affected by the disaster." Although the location of the inventory has not been disclosed in the reports this time, it is a point that must be confirmed from the perspective of BCP design.

12. Why do plants stop even when prepared? The structural limitations of BCP

Based on the facts so far, let's move on to a more general question. Why do plants stop even for companies that have prepared seriously? I will list five structural reasons.

First, there is the problem that countermeasures are optimized for past damage patterns. Fixing pipes that fell last time and securing molds that flew out last time is correct as a learning process, but the next earthquake will not necessarily destroy the same places as the last one. The cycle, direction, and acceleration bandwidth are all different. As long as the list of areas to take countermeasures is created from 'past damage records,' there will always be omissions in the list.

Second, there is the problem that utility equipment is easily left out of the priority list for countermeasures. This time, it was reported that equipment used to cool production facilities could not be operated. Cooling systems, compressed air, pure water, exhaust, and power supplies are not 'production equipment,' so their priority is lowered in capital expenditure requests for seismic reinforcement. However, if any one of these stops, the production equipment will not run. Whether the equipment list is created based on 'what is necessary for production' or 'asset ledgers' changes how omissions occur in countermeasures.

Third, there is the problem that operational decisions are halted by the uncertainty of 'not knowing if it will work.' This time, stakeholders said they 'had not yet been able to test' whether the production equipment would operate. If it is confirmed to be broken, it can be fixed. However, a state of not knowing whether it is broken or not stops decision-making. Especially in automotive parts where quality assurance is strict, the choice to resume production with equipment whose precision cannot be guaranteed cannot be taken. There are very many cases where this 'verification time' is not incorporated into BCP plans.

Fourth, there is the problem that the bottleneck is determined by the single weakest point. Factory utilization rates are not determined by averages. One out of three buildings, one part number out of countless molds, one machine out of many pieces of equipment. If this stops, that line stops. If investment is distributed evenly, this weakest point remains. BCP investment should be allocated to identifying and reinforcing the weakest points, not to raising the average.

Fifth, there is the problem that geographical concentration arises from economic rationality. If low-unit-price parts like door checks are produced in a decentralized manner, mold investment overlaps, logistics costs increase, and profitability worsens. Therefore, they are consolidated. Semiconductors are the same. TSMC chose Kumamoto, and Sony, Renesas, and Tokyo Electron gathered in the same region because the agglomeration effects of human resources, infrastructure, and supplier networks are at work. Agglomeration increases productivity, but at the same time, it increases geographical risk. These two are two sides of the same coin, and one cannot unilaterally deny either.

To summarize the above, the conclusion is that the realistic goal of BCP should not be 'not to stop,' but 'not to let it spread even if it stops' and 'to return in a short time even if it stops.' The Kumamoto case this time shows that this direction was correct. At the same time, it also shows that there are still parts within that direction that are insufficient.

13. Five perspectives that management should learn

For those involved in manufacturing management, or those who want to inspect their own supply chains, I will organize the practical issues that can be extracted from this case.

The first is to look at risk by part number, not by part unit. What fell this time was 'a mold for one part number.' Supply chain visualization is often done by 'supplier unit.' However, what actually stops the line is the part number. Even in the same factory of the same supplier, molds differ by part number, and substitutability differs. It is worth checking whether the granularity of visualization has been reduced to the part number level.

The second is to design inventory days and recovery target time as a set. As mentioned in the previous section, the figure of 5 days of inventory can only be evaluated in conjunction with the premise of how many days it will take for alternative production to start up. It is necessary to create a state where the inventory decided by the finance department and the recovery plan held by the production engineering department are checked against each other on the same table.

The third is to explicitly include utility equipment in the scope of seismic countermeasures. Cooling systems, compressors, pure water equipment, and power receiving and transforming equipment. Just by adding these to the list of 'production equipment,' omissions in countermeasures will be greatly reduced. The report this time that 'equipment to cool production facilities could not be operated' most clearly shows this point.

The fourth is to incorporate the 'verification time' after a disaster into the plan. Time to confirm that equipment is not broken, time to recalibrate precision, time to confirm quality with prototypes. If the recovery plan assumes this time is zero, the actual recovery will take two to three times longer than planned. The process of 'testing whether it works' can actually become the critical path for recovery.

The fifth is to confirm the physical location of the inventory. If the inventory is inside a building in the disaster area, it cannot be retrieved due to entry restrictions. The countermeasure of holding inventory only functions under the condition of 'holding it in a place that will not be affected by the disaster.' Placing inventory in geographically distant locations is costly, but inventory that does not pay that cost becomes numbers on paper in an emergency.

None of these five require huge investments. Changing the granularity of visualization, checking numbers across departments, broadening the definition of the countermeasure list, adding one process to the plan, reviewing the storage location of inventory. All are issues of design and operation. The direction of polishing the design of containment, rather than pouring infinite costs into prevention, is consistent here as well.

14. How investors should read this event

From an investor's perspective, I will organize how this event should be evaluated. It is effective to think by dividing it into three time axes.

From a short-term perspective, the focus is on the extent of the impact on production volume. The stoppages this time were Toyota's three Fukuoka factories from the evening of July 29 to the 31st, some lines at Nissan's two Fukuoka factories from the 30th to the 31st, and Daihatsu Kyushu's Kurume factory until the 31st, with the Oita factory already having resumed operations. In terms of days, it is on a scale of a few days, which is a different level from the three-week scale of 10 years ago.

However, it is necessary to be aware that the situation after August 3 is undecided. Whether the stoppage exceeds one week is the turning point that affects quarterly production volume. Since automakers usually perform catch-up production in the remaining period until the end of the term, a stoppage of a few days hardly affects the annual volume. Whether catch-up is possible is determined by the length of the stoppage period, the remaining days until the end of the term, and the room for holiday operations.

From a medium-term perspective, the vehicle model composition of the Kyushu base is important. The Lexus from the Miyata plant is for export to China and elsewhere, and the Nissan Rogue is for North America. Both are models that contribute significantly to profit per vehicle, and there is a possibility that the sensitivity to profit is higher than the volume ratio. Reading the figure of 'a little over 10% of domestic production' as a 10% impact on profit as is is a mistake. It is necessary to check the mix of vehicle models.

Also, trends on the semiconductor side are a medium-term issue. JASM confirmed structural safety and resumed operations in stages. The resumption date for Sony's Kikuyo plant was left undecided. The meaning of risk changes depending on whether semiconductors are in a phase where inventory is being accumulated or a phase where it is depleted. Especially if there are items with thin inventory, such as image sensors and microcontrollers for automotive use, it will have a delayed impact on automotive production plans. At the financial results briefing, specific questions about the status of parts inventory should be raised.

From a long-term perspective, there is the issue of how to incorporate the risk premium of industrial clusters. Over the past 10 years, the semiconductor cluster in Kumamoto has grown to a scale of 991.9 billion yen in shipments and 16,892 employees. There have been 70 location agreements in total. While this concentration is a source of productivity and competitiveness, it is also a concentration of geographical risk. The reports that no large-scale equipment damage has been confirmed are a source of relief, but we should have separate assumptions for cases where aftershocks continue or a larger earthquake occurs.

As a practical matter for investment decisions, it is worth confirming the following three points with companies: How many part numbers depend on a single site? Are the days of inventory consistent with the lead time for starting up alternative production? For geographically concentrated sites, to what extent has the effectiveness of alternative production been verified? Companies that can answer these three questions specifically are considered to have high resilience in times of crisis.

Note that this is not investment advice. The evaluation of individual stocks is an area where you should directly check financial statements and company disclosures. What I have presented here is an organization of points for interpreting events.

15. Weekend work is the climax: What is happening at the recovery site

Moving away from facts and analysis, let's return to the story on the ground.

The reporter who wrote the Nikkei article is from the Jonan district of Minami-ku, Kumamoto City, where this factory is located. It is written that their parents' home was also hit by the disaster again, but the damage to the house was minor. It is said that 'Crea,' or Aeon Mall Kumamoto, where the explosion occurred, was also a place they frequented. The reporter notes that their elementary, junior high, and high school classmates, as well as seniors and juniors, are working hard on reconstruction locally.

Ten years ago, the reporter, who was dispatched to the site from Nagoya, saw hundreds of support troops gathered from all over the country, wiping away sweat. It is said that support teams have gathered from all over the country this time as well.

At the entrance to the factory office, a message board sent during the disaster 10 years ago is still displayed. It features the words 'Don't give up, Kumamoto' and an image of Kumamon. The Aisin Heritage Museum displays the helmet used during the 2016 Kumamoto earthquake. The 10 years have been a time not only for weathering the countermeasures but also for preserving memories as a system.

The words of those involved, 'Weekend work is the climax,' may be the most significant sentence in this article. Five days of inventory is a tool to buy time. However, it is people who do the work during that time. From the weekend of July 31 to August 2, at the Kumamoto factory, the work of testing whether equipment works, repairing pipes, putting back fallen molds, and starting up cooling devices will continue. There are people who have experienced a magnitude 7 earthquake and whose families have also been affected, yet they still go to the factory.

Discussions about supply chains tend to drift toward abstract terms like days of inventory and risk diversification. However, what supports the end of that is the hands of the people holding tools in the disaster area. The report that all employees on duty were safe this time has great value in itself. Equipment can be repaired, but people cannot.

The Nikkei reporter concludes the article this way: 'Nansamo, gamadase Kumamoto.' In any case, do your best, Kumamoto. And, reconstruction is definitely possible.

16. To avoid repeating the same question in 10 years

Finally, I will summarize the conclusions to be drawn from this event.

First, in this Kumamoto earthquake, the Aisin Kyushu factory was hit by a disaster for the third time, and there is no prospect for resuming operations. Although 10-year countermeasures such as seismic equipment for ceiling-suspended pipes and stoppers for mold shelves had been implemented, some pipes burst, one mold for a part number fell from the shelf, and the device for cooling production equipment became inoperable.

Second, due to the shutdown of this factory, Toyota suspended three factories in Fukuoka Prefecture until the 31st, Nissan suspended some lines at two factories in Fukuoka Prefecture from the 30th to the 31st, and Daihatsu suspended factories in Oita and Kurume until the 31st. Operations after August 3 are undecided.

Third, however, the impact is expected to remain within the Kyushu region. Ten years ago, 26 out of 30 domestic lines stopped, resulting in a production delay of about 80,000 units. This time, the factories in Aichi Prefecture have not stopped. What created this difference are the 10-year countermeasures of 5 days of inventory, production of the same product at multiple locations, and the dualization of logistics networks.

Fourth, therefore, the assessment that '10-year countermeasures also fell short' is half correct and half inaccurate. It failed to keep the factory from stopping. It succeeded in stopping the spread to the rest of the country. If the realistic goal of BCP is containment rather than prevention, the investment of the last 10 years was in the right direction.

Fifth, there are still remaining issues. Alignment of inventory days and recovery speed, missed countermeasures for utility equipment, estimation of verification time after a disaster, risk visualization by part number, and the physical location of inventory. These are areas that can be improved by design rather than investment amount.

Sixth, in the last 10 years, Kumamoto has become a semiconductor cluster. It has a scale of 991.9 billion yen in shipments, 16,892 employees, and 70 location agreements. The structure where both automotive mechanical parts and automotive semiconductors are placed under the same regional risk did not exist 10 years ago. Risk diversification design needs to be evaluated simultaneously on two axes: diversification by part category and geographical diversification.

And finally, this is a story of repetition. The Nikkei reporter begins the article with the sentence, 'The unthinkable has happened again.' The same factory was hit by a disaster three times, and the same industry stopped twice. There is no reason to think that a fourth time will not come in 10 years.

We need to reframe the question. Instead of 'How can we prevent a shutdown?', we should ask, 'When a shutdown occurs, how quickly and within how narrow a scope can we contain it?' Companies that can answer this question with concrete numbers will be the ones to survive the next decade.

Restoration work is currently underway at a rapid pace at the factories in Kumamoto. First and foremost, I would like to wish for the safety of the people working there and for the recovery of the region.

References and Sources

• Nihon Keizai Shimbun, 'Toyota and Nissan halt production due to Aisin Kumamoto disaster; 10 years of countermeasures fall short' (July 30, 2026, Digital Edition, Reporter: Ryo Asayama)

• Prime Minister's Office of Japan, 'July 28, 2026, Chief Cabinet Secretary Press Conference' https://www.kantei.go.jp/jp/105/statement/2026/0728kaiken.html

• Bloomberg, 'Magnitude 7 earthquake in Kumamoto' related reports (July 29, 2026) https://www.bloomberg.com/jp/news/articles/2026-07-29/TIWRD9KJH6V500

• FNN Prime Online, 'Explosion at Aeon Mall Kumamoto' related reports (July 2026) https://www.fnn.jp/articles/-/1083802

• FNN Prime Online, 'Aisin Kyushu operations halted' related reports (July 2026) https://www.fnn.jp/articles/-/1084130

• Jiji Press, 'Impact of the Kumamoto earthquake on corporate activities' (July 29, 2026) https://www.jiji.com/jc/article?k=2026072900547&g=eco

• Newswitch / Nikkan Jidosha Shimbun, 'Toyota Motor Kyushu 3-plant operations halted' related reports (July 2026) https://www.netdenjd.com/archives/685083

• nippon.com'Kumamoto earthquake, corporate damage status' (July 28, 2026) https://www.nippon.com/ja/news/yjj2026072801023/

• Reuters, 'TSMC says Japan plant gradually resuming operations after earthquake' (July 28, 2026) https://www.reuters.com/world/asia-pacific/tsmc-says-japan-plant-gradually-resuming-operations-after-earthquake-2026-07-28/

• Toyo Keizai Online, 'Kumamoto semiconductor plant, operational status after the earthquake' (July 2026) https://toyokeizai.net/articles/-/953056

• Kumamoto Nichinichi Shimbun, 'Operational status of semiconductor companies in the prefecture' (July 29, 2026) https://topics.smt.docomo.ne.jp/article/kumanichi/region/kumanichi-dr2019448

• Aisin Seiki, 'Regarding the impact of the Kumamoto earthquake on production activities' (April 28, 2016) https://www.aisin.com/jp/news/2016/000902.html

• Toyota Motor Corporation, "Regarding the Status of Vehicle Production Following the Kumamoto Earthquake" (April 17, 2016) https://global.toyota/jp/detail/11811462

• Toyo Keizai Online, "The Shock of Toyota Suspending All Domestic Plants" (April 2016) https://toyokeizai.net/articles/-/114251

• J-CAST News, "The Full Story of Toyota's Production Suspension During the Kumamoto Earthquake" (May 11, 2016) https://www.j-cast.com/2016/05/11266357.html

• International Research Institute of Disaster Science, Tohoku University, "Report on Corporate BCP Response to the Kumamoto Earthquake" (April 19, 2016; May 20, 2016) https://irides.tohoku.ac.jp/media/files/earthquake/eq/20160419_kumamotoeq_bcp_maruya.pdf

• Kyushu Bureau of Economy, Trade and Industry, "Current Status of the Automotive Industry in Kyushu" https://www.kyushu.meti.go.jp/jirei/kokusai/investment/2024/240409_1_6.pdf

• Kyushu Bureau of Economy, Trade and Industry, "Current Status of the Kyushu Economy 2025" https://www.kyushu.meti.go.jp/keiki/chosa/genjyo/kyushu-genjo_2025.pdf

• Regional Economy Research Institute, Inc., "Trends in the Semiconductor-Related Industry in Kumamoto Prefecture" (March 2026) https://www.reri.or.jp/wp/wp-content/uploads/2026/03/db745523ac68fac5631cbc878f33dea0.pdf

• Kumamoto Prefecture, "Industrial Statistics and Economic Structure Survey: Overview of the Manufacturing Industry" https://www.pref.kumamoto.jp/uploaded/attachment/299046.pdf

• Policy Research Institute, Ministry of Finance, "Semiconductor Industry Agglomeration and the Regional Economy" (January 2025) https://www.mof.go.jp/pri/research/conference/fy2024/naigai202501_2.pdf

• Toyota Motor Kyushu Recruitment Information (Export Ratio of Produced Vehicles) https://www.toyota-kyushu.com/recruit/

*This article is written based on information published and reported as of July 30, 2026. Damage status, production suspension periods, and recovery outlooks may change in the future. Please check the official announcements of each company for the latest information. Furthermore, this article does not recommend the buying or selling of any specific financial products.