Toyota and Suppliers to Invest ¥1 Trillion a Year from 2028: How to Pick What to Replace

News, 13 November 2026. Every news figure below comes from a linked source, and the full list is at the end.

What Toyota said on 11 November 2026

Toyota described its plans at an event at its North American R&D headquarters in Saline, Michigan, and in a release on its US newsroom. Manufacturing was one of five areas, under the heading of physical AI and robotics.

The release says Toyota and its suppliers expect to invest about ¥1 trillion (about $6.3 billion) a year from 2028 onward to renew aging production equipment. That investment will support the deployment of about 150,000 robots across Toyota and Toyota Group companies and 250,000 at suppliers.

Toyota sets out three levels of factory automation:

• FA 1.0: conventional robots for predefined, repetitive tasks, to reduce physical strain.
• FA 2.0: more automation, with production and logistics synchronized end to end.
• FA 3.0: people and robots working side by side.

The order matters. Toyota says the approach starts with team members simplifying processes through the Toyota Production System, then finding where automation and robotics can help them. It cites the expertise of some 18,000 Takumi, its highly skilled workers worldwide, and says physical AI and robotics will learn from its expert Takumi. Toyota also describes a Factory Control Platform that connects equipment, logistics and the supply chain.

The numbers behind the news

The table puts Toyota's figures next to the robot market they will land in. The first six rows are from Toyota's release; the last five are from the International Federation of Robotics (IFR).

Sources: Toyota release, IFR on Japan's car industry and IFR World Robotics 2025.

Two things stand out. First, suppliers carry the larger share of the robots: 250,000 of 400,000, or 62.5%. Toyota's plan is as much a supplier plan as a Toyota one. Second, the scale is large next to today's market. 400,000 robots is about 31 times the 13,000 robots Japan's whole car industry installed in 2024.

What the numbers do not say matters just as much. The release gives no end date for the robot rollout, no split by country or plant, and no split of the yen between robots and other equipment. It does not say how many existing machines will be retired. So treat the 400,000 as a direction of travel, not a delivery schedule, and do not divide the yen by the robots to get a price per robot.

Background: why this lands now

Japan's car industry was already buying more robots before this announcement. The IFR reported that it installed about 13,000 industrial robots in 2024, up 11% on 2023 and the highest level in five years. Its robot density was 1,531 robots per 10,000 employees in 2023, fourth in the world.

The wider Japanese market moved the other way. Installations across all industries fell about 4% to 44,500 units in 2024, though Japan stayed the second-largest market after China. Worldwide, 542,000 robots were installed in 2024 and 4,664,000 were in operation. The IFR also says Japan makes 38% of the world's robots. Demand has risen since: the Japan Robot Association's July to September 2026 results show member orders of 67,003 units, up 36.1% on a year earlier, worth ¥325.4 billion, up 46.6%. Domestic shipments to car and auto parts makers were 2,533 units, up 15.5%, or 27.6% of all domestic shipments. Machine tools tell the same story: the Japan Machine Tool Builders' Association's preliminary October 2026 figures put domestic orders at ¥52.3 billion, 46.5% above October 2025.

Against that background, Toyota's message is less about buying robots and more about renewing old equipment in a set order. Its own description of the Toyota Production System says that before starting with a machine, you should "first try doing it thoroughly by hand". The release pairs the money with three automation levels, its Toyota Production System and the knowledge of some 18,000 Takumi. Its Factory Control Platform is meant to connect equipment, logistics and the supply chain.

Why age is a poor guide to what to replace

Few plants will spend at Toyota's scale. Almost every plant, though, has the same problem in miniature: a fleet of older machines, a capital budget that covers only some of them, and pressure to automate.

The usual shortcut is to replace the oldest machine first. That often picks the wrong one. A 22-year-old press that runs all day can lose less output than a 9-year-old filler that stops every few minutes.

Toyota's sequence is the useful part for everyone else: simplify first, then automate. A robot added to a messy process runs the mess faster. We have no view of how Toyota will pick its machines and make no guess about it.

Rank by lost output, not by age: a worked example

Example data, not Toyota's. Two machines each have 100 planned hours a week.

Lost hours here are planned hours times the share OEE does not capture: 100 x 19% = 19, and 100 x 42% = 42. The newer machine loses more than twice as much. Its two biggest losses, changeovers and short stops, are process problems that a new machine would not fix by itself.

Have an old fleet and a short capital budget? Book a demo and see machines ranked by lost hours, not age.

What it means for plant, production and maintenance managers

Plant managers: capex timing and customer approvals

If your plant supplies Toyota or another carmaker, the 250,000 supplier robots are the line to read twice. A renewal wave across many suppliers can stretch integrator and robot lead times, so a cell you want running in 2028 may need an order and a floor plan well before then. Decide early which lines are candidates and which are not. In automotive, changing the equipment that makes an approved part usually means telling the customer and repeating some part approval work. Put that approval time into the project plan, not after it. Hiring is the other decision: a robot cell needs fewer operators per shift but more technicians who can recover a fault, re-teach a path and read a fault log. If you have one such person on day shift only, nights become the weak point the day the cell goes live.

Production managers: ramp-up, schedules and targets

A new cell rarely runs at its rated speed in its first weeks. Build a bank of parts before cutover so customer shipments do not depend on a cell that is still learning. Where floor space allows, keep the old machine able to run for a few weeks after the new one starts. Set OEE targets per line by ramp week, not one target from day one, and write down what the old machine achieved so the comparison is fair. Rethink staffing per shift: an operator who tended one manual station may now load two or three cells, so the shift plan and the break cover change. Changeovers deserve their own check. A robot that is fast in cycle but slow to change tooling or programs can lose its gain on a high-mix line. Our new line ramp-up plan covers the week-by-week targets.

Maintenance managers: the old fleet and the new one

The risky period is before 2028, not after. Machines that are due to be replaced tempt teams to skip preventive work, and run-to-failure on an old press or welder can stop a line for weeks if parts are no longer made. Keep PM on those assets until the day they leave, and check now which spares are already obsolete. For the new robots, plan the spares list during the purchase, not after the first failure: servo motors, cables and dress packs, teach pendants and grippers are typical items. Write the robot maker's PM schedule into your maintenance system on day one. Commissioning is a maintenance job too: agree acceptance tests with the integrator, record a baseline of cycle time and faults, and make the first warranty claims with data rather than memory.

Three steps before you replace a machine

1. Rank machines by lost output. Measure OEE per machine for four to eight weeks, then sort by lost hours, not by age or by how loud the complaints are. Our guide to the replace vs repair decision adds the cost side.
2. Split the top losses before you buy. If the biggest losses are changeovers, waiting or short stops, simplify the process first. That is Toyota's own order, and it is often cheaper than new equipment.
3. Keep the before for the after. Use the same stop reasons and the same rated speed on the old machine and the new one. Then a robot cell has to prove its gain in output, not only in the vendor's slides.

What to do this quarter

• Freeze a baseline now. Record six weeks of OEE and stop reasons on every machine that might be renewed by 2028. Owner: production manager.
• Tag each candidate by its top loss. If more than half of a machine's lost hours are breakdowns, it is a renewal candidate; if they are changeovers or short stops, fix the process first.
• Price the old fleet's risk. List machines with obsolete spares and estimate the days of downtime if each one fails. Owner: maintenance manager.
• Map customer approvals. List every approved part whose equipment would change and how long each customer's change process takes. Owner: quality manager.
• Ask for lead times in writing. Get two integrator quotes with delivery and commissioning dates, so the capex calendar rests on dates, not hopes.
• Count your robot skills per shift. Target at least one technician per shift who can recover a robot fault; plan training for any shift without one.
• Write ramp terms into the order. State the OEE or cycle time the new cell must reach and by which week, measured the same way as the baseline.

Where OEE and light MES help

OEE per machine turns the replacement list into a ranking by lost output. Availability, performance and quality show whether a machine fails, runs slowly or makes scrap, and each one points to a different fix.

A light MES keeps the stop reasons and shift reports that explain the number. Operators add the reason on a tablet or phone at the machine, so short stops and changeovers are visible, not lost in an average.

Older machines often have no PLC data to read. A Fabrico IoT device with sensors can record their run and stop time, and on manual stations an AI camera (an add-on) can count output, so the old fleet is measured like the new one. See OEE on legacy machines.

Which OEE factor does equipment renewal touch most? On paper, availability: old machines break down and new ones should not. In practice, the first months after a robot cell goes in often hit performance and availability at the same time, from slow cycles, re-teaching and small faults. That is why the before and after must be measured the same way.

Example data. An old welding station with 100 planned hours a week runs at 85% availability, 90% performance and 98% quality: 0.85 x 0.90 x 0.98 = 75.0% OEE. In its first week, the new robot cell runs at 70% availability, 80% performance and 97% quality: 0.70 x 0.80 x 0.97 = 54.3% OEE. That is 20.7 points lower. If it reaches 92%, 95% and 99% by week eight, OEE is 0.92 x 0.95 x 0.99 = 86.5%, or 11.5 points above the old station. Judge the cell in week eight, not week one, and only on the same stop reasons and rated speed.

In money (Example data). Assume an ideal rate of 30 parts an hour and a contribution of $15 per good part. Good parts a week are OEE x 100 hours x 30 parts. The old station makes 75.0% x 3,000 = 2,250 parts, or $33,750 a week. Week one of the robot cell makes 54.3% x 3,000 = 1,629 parts, $9,315 less (621 x $15). By week eight it makes 86.5% x 3,000 = 2,595 parts, $5,175 a week more than the old station (345 x $15), or $269,100 over 52 weeks. That annual figure, not the robot's price alone, is what the capex case should rest on.

Choosing a tool to rank your fleet? Our guide to OEE monitoring software covers the options.

How Fabrico does it

Fabrico is a cloud OEE and light MES platform. It shows OEE and losses by machine, line and shift from the PLC or a Fabrico IoT device, with the reasons operators add on a tablet or phone. Where there is no signal, AI cameras (an add-on) count output. We have no link to Toyota; the figures above are from its release.

Deciding which old machines to replace first? Book a demo.

What people ask next

How much will Toyota invest in new production equipment?

Toyota says that it and its suppliers expect to invest about ¥1 trillion, about $6.3 billion, a year from 2028 onward to renew aging production equipment.

How many robots does Toyota plan to deploy?

The investment will support the deployment of about 150,000 robots across Toyota and Toyota Group companies and 250,000 at suppliers, according to Toyota's release of 11 November 2026.

Which old machine should be replaced first?

Rank machines by lost output, not by age. Measure OEE per machine for a few weeks, sort by lost hours, and check whether the top losses are process problems, such as changeovers or short stops, that a new machine would not fix. See our guide to the replace vs repair decision.

How many robots in total does Toyota's plan cover?

About 400,000: some 150,000 across Toyota and Toyota Group companies plus 250,000 at suppliers, according to Toyota's release of 11 November 2026. The release does not give an end date for the rollout or a split by country or plant.

What are FA 1.0, FA 2.0 and FA 3.0?

They are Toyota's three levels of factory automation. FA 1.0 uses conventional robots for predefined, repetitive tasks to reduce physical strain. FA 2.0 adds automation and synchronizes production and logistics end to end. FA 3.0 is people and robots working side by side.

How many robots does Japan's car industry install a year?

About 13,000 in 2024, up 11% on 2023 and the highest in five years, according to the International Federation of Robotics. Its robot density was 1,531 robots per 10,000 employees in 2023, fourth in the world.

Sources

All sources accessed on 14 November 2026. Dates are the publication dates shown on each page.

• Toyota, Toyota Powers What's Next in Mobility and Manufacturing, 11 November 2026
• International Federation of Robotics, Japan's Car Industry has Highest Robot Installations in Five Years, 15 July 2025
• International Federation of Robotics, World Robotics 2025 report: global robot demand in factories doubles over 10 years, 25 September 2025
• Japan Robot Association, July to September 2026 results: production and shipments of manipulators and robots, 19 October 2026 (file date)
• Japan Machine Tool Builders' Association, preliminary machine tool orders, October 2026 (PDF, Japanese), 11 November 2026
• Toyota Motor Corporation, Toyota Production System, accessed 14 November 2026