Toyota Says Factory Automation Could Require 400,000 Robots From 2028

The 400,000 Number Is a Factory-Automation Estimate, Not a Humanoid Order

The easiest way to misread Toyota's disclosure is to treat every robot in the estimate as a general-purpose humanoid. Reuters explicitly says otherwise. The 400,000 figure covers conventional and newer robot types, replacement equipment and incremental installations, and a broad manufacturing network that extends beyond Toyota Motor itself to group companies and major suppliers.

That makes the number less sensational, but arguably more useful. Humanoids are entering factories that already contain enormous amounts of fixed automation. Their economic case will depend on whether they can solve the awkward remainder: mobile material handling, variable manipulation, machine tending, inspection, and human-designed workflows that are expensive to automate with a custom cell.

Toyota told investors that the potential modernization effort spans industrial robots, automated logistics, and future human-robot collaboration. Those categories overlap. A future factory may use fixed arms for repetitive precision work, autonomous mobile robots for transport, specialized manipulators for machine interfaces, and humanoid or human-shaped systems where the environment and task variability make generality valuable.

Toyota Already Has a Factory Automation Baseline

The new estimate is easier to understand in the context of Toyota's own manufacturing reports. In its 2025 Integrated Report, Toyota described a piston assembly line at its Kamigo Plant that began running with robots in January 2025. The company said that line achieved full automation, reducing the operator count from three to zero, while noting that no other line at Kamigo had reached the same level.

That disclosure is important because it shows how Toyota thinks about automation. The company did not present a simple story of replacing people with robots everywhere. It described a long history of using manual work, jigs, process design, and selective automation, then applying robots where technology and maintainability finally made the trade worthwhile.

Toyota is also developing what it calls a Co-Creation Line, intended to bring more automation into engine manufacturing while preserving the production knowledge that workers use to improve processes. One example in the report is a robot designed to set bolts on a timing-chain cover, a task that became more practical to automate as camera technology improved.

The 2028 estimate therefore looks less like a sudden robot shopping spree and more like an acceleration of a factory-renewal program. Aging equipment, labor shortages, logistics automation, better machine vision, and physical AI can all push the same capital cycle in the same direction.

Where Humanoids and Physical AI Fit

Toyota has several robotics paths that could feed into that modernization. Biped.News previously covered a seven-robot Agility Digit deployment at Toyota Motor Manufacturing Canada, a narrow material-handling program that illustrated one reason humanoids appeal to automakers: they can move through facilities built around human dimensions without requiring every workstation to be redesigned.

Toyota's internal research points at a different part of the problem. In March, the company's Frontier Research Center introduced ELEY, a human-shaped mobile manipulator designed around compliant, backdrivable joints and learning from human movement. Toyota said ELEY is intended to work toward production-site use, while candidly listing reliability during long operation, repeatability, and data infrastructure as areas that are not yet at leading global levels.

That gap between research and deployment is exactly why the humanoid share of the 400,000 matters. A conventional welding robot has a mature safety case, predictable cycle time, and decades of integration knowledge. A general-purpose physical-AI robot promises flexibility, but it must still prove uptime, recovery behavior, safety around people, intervention rate, and cost per useful task.

Toyota's scale can make those proof points unusually consequential. A system that saves a few minutes on one task is interesting in a pilot. A system that can be standardized across dozens of plants and suppliers can change the economics of an entire robot category.

Why 2028 Keeps Showing Up

Toyota's potential spending starts from 2028, a date that also appears in other automotive humanoid plans. Hyundai has said it intends to begin deploying Boston Dynamics Atlas humanoids at its U.S. operations from 2028. That does not mean the industry has agreed on a universal launch year. It does suggest that automakers are planning around the time required to finish product development, safety validation, plant integration, procurement, and worker training.

For robot vendors, those lead times matter. Winning a factory program in 2028 may require qualifying hardware and software well before then. Automotive production systems are designed around uptime and controlled change. A robot that performs a task in a demo must still survive shift schedules, maintenance procedures, parts variation, network outages, safety reviews, and the occasional strange object that was never in the training set.

The larger implication is that physical AI is entering the same procurement world as established industrial automation. Buyers will care less about a robot's best video and more about mean time between incidents, recovery time, fleet-management tools, safety certification, integration effort, and total cost per completed unit of work.

The Labor Story Is More Complicated Than Replacement

Reuters tied the automation push partly to aging infrastructure and labor shortages. Those pressures can favor robots even when a factory is not trying to reduce total employment. A plant may have trouble staffing repetitive, hazardous, or ergonomically difficult jobs while still needing more people in maintenance, quality, engineering, and higher-value production roles.

Toyota's own factory reporting emphasizes that distinction. Its Kamigo discussion frames automation as a way to keep plants competitive and shift workers toward more valuable tasks, while preserving the production knowledge that allows processes to improve. Whether that balance holds at the scale implied by 400,000 machines is an open question, not something the investor estimate answers.

The mix of robots will matter too. Replacing an aging fixed arm with a newer fixed arm changes a job differently than introducing a mobile robot that can move between stations. A humanoid capable of learning multiple tasks could eventually change staffing models more broadly, but today's systems still require integration, supervision, maintenance, and carefully bounded workflows.

What to Watch Next

The first number Biped.News wants is the humanoid share. Toyota has now supplied a total robot requirement, but not the portion expected to come from human-shaped or general-purpose machines. Even a low-single-digit percentage of 400,000 would represent thousands of units, while a much smaller share would show that conventional automation remains the dominant answer.

The second is vendor allocation. Toyota already touches multiple robotics ecosystems through internal R&D, manufacturing partners, startups, and supplier relationships. A broad modernization program does not imply a single winner. Different plants and tasks are likely to use different machines, and the most valuable vendor may be the one that integrates cleanly with Toyota's production system rather than the one with the most humanlike hardware.

The third is operating evidence. Watch for site names, fleet counts, paid deployments, uptime, intervention rates, safety certifications, and task-level productivity. Those metrics will show whether physical AI is becoming a meaningful slice of Toyota's automation stack or remains a small experimental layer beside mature industrial robotics.

Frequently Asked Questions

Is Toyota buying 400,000 humanoid robots?

No. Toyota's estimate covers roughly 400,000 robots of multiple types across Toyota, group companies, and major suppliers. Reuters says it includes humanoid and non-humanoid robots, replacements for existing machines, and new installations.

Has Toyota committed to spending $6.4 billion per year?

Not yet. Toyota told investors that factory modernization could cost about ¥1 trillion, or $6.4 billion, annually from 2028, but did not say the spending would definitely occur or how many years it would continue.

Where would the robots be used?

Toyota described a broad automation effort spanning industrial robots, automated logistics, and future human-robot collaboration on factory floors. The company has not published a site-by-site allocation for the 400,000 estimate.

What role could humanoids play?

Humanoids are most relevant where mobility and manipulation are needed in spaces designed for people and where fixed automation is too rigid or expensive. Toyota has not disclosed how many humanoids it expects to use, so any unit forecast is still speculation.