What Toyota jidoka, Tesla and BMW teach Indian manufacturers about automation
SRF Capital Studio Research DeskFunding Intelligence, SRF Capital StudioThree carmakers came at automation from different directions and arrived at rules small enough for a 150-worker plant.
Summary
- Toyota's jidoka builds machines that stop themselves when something goes wrong, so a person fixes the cause before more bad parts are made.
- Tesla automated its Model 3 line so heavily that output fell well short of target in 2018, and Elon Musk called the excess automation his own mistake.
- BMW measured one humanoid robot on one job for ten months, then took the lessons to a second plant. An Indian supplier can copy all three at its own scale.
Picture a plant in Coimbatore where 150 people make brackets for a carmaker. Its owner reads about Toyota, Tesla and BMW and sees giants with nothing to teach him. Yet the useful part of their stories costs nothing: the order in which they automated, and what they learned when it went wrong.
What does jidoka mean at Toyota?
Toyota translates jidoka as automation with a human touch. It traces the idea to Sakichi Toyoda's automatic loom, which could judge when something had gone wrong and stop itself.
In Toyota's plants, a machine that detects a fault stops on its own, and any operator can halt the line by pulling a cord. Bad parts stop flowing at once. A person then finds the cause and fixes it, so the fault does not come back.
Jidoka is one of two pillars of the Toyota Production System. The other is just-in-time: making only what the customer needs, when it is needed, in the amount needed.
Read that way, a machine's first job is to expose a problem and call a person. Saving labour comes later, as a result of a process that runs cleanly. A plant that automates an unstable step only makes its defects faster.
A plant that automates an unstable step only makes its defects faster.
What went wrong when Tesla automated the Model 3?
Tesla went the other way. It built its Model 3 line with a high degree of automation and aimed for 2,500 cars a week by the end of March 2018. In the last week of that quarter it made 2,020, Fortune reported.
On 13 April 2018 Elon Musk wrote on Twitter: "Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated." He told CBS News the plant had a "crazy, complex network of conveyor belts" that was not working, so Tesla removed it.
Our reading is that each automated step was a new place for the line to stop, and the process was too new to run smoothly. People absorb small upsets that a rigid machine cannot, and Tesla needed that while it learned.
How did BMW test humanoid robots before scaling up?
BMW took the slow road. In 2024 it trialled Figure 02, a humanoid from the Californian company Figure, at its Spartanburg plant in South Carolina. The robot placed sheet metal parts into fixtures for welding.
In 2025 the robot moved into regular production. Working ten-hour shifts, Monday to Friday, it handled more than 90,000 parts in about 1,250 hours. Over ten months it helped build more than 30,000 BMW X3s, the company said in February 2026.
That works out to about 72 parts an hour, one every 50 seconds. A person can keep that pace, so the trial's value lay in what it proved about reliability on a real line.
Then BMW named a second site. At Leipzig in Germany, it said, Hexagon's AEON humanoid would work on high-voltage battery assembly and component manufacturing, with a pilot phase from summer 2026. BMW also set up a centre to pool what it learns from such robots.
What should an Indian plant copy from all three?
- Stabilise the job first. Find out why defects and stoppages happen and fix the causes, then automate the step that is left. That is Toyota's order and Tesla's lesson.
- Make machines stop and call for help. A sensor that halts a press when a part sits wrong is cheap next to a robot, and it stops scrap at the first piece.
- Automate one step at a time. Pick the most repetitive, tiring or risky job and leave the rest to people until it runs well. A dark factory, one that runs with no people on the floor, is a poor first goal.
- Set the numbers before the trial. Fix the target for parts an hour, defects and stoppages, run for a set period, then decide. BMW counted hours, parts and cars.
- Keep people where judgment matters. All three companies still lean on people to spot and solve problems on the line.
The money side needs the same discipline. Our guide to the robot payback period shows how to test a cell before you buy, and what a cobot costs in India prices the usual first step. Where India stands overall is in our note on robot density.
If the process itself needs work before any machine arrives, that is what our process optimisation team does for manufacturers.
Frequently asked questions
What is jidoka in Toyota's production system?
Toyota's term for automation with a human touch: machines that detect a fault and stop themselves, so a person can fix the cause. It is one of the system's two pillars, with just-in-time.
What did Elon Musk mean by 'humans are underrated'?
In April 2018, with Model 3 output behind target, Musk admitted that excessive automation at Tesla had been a mistake, and his own. People on the line, he meant, had been undervalued.
What was Tesla's production hell?
The phrase used for Tesla's struggle to build the Model 3 in volume in 2017 and 2018. At the end of March 2018 it made 2,020 cars in a week against a target of 2,500.
How does BMW use robots in its factories?
Beyond its conventional robots, BMW is piloting humanoids. A Figure 02 worked at Spartanburg in 2025 and helped build more than 30,000 X3s. BMW planned a pilot of Hexagon's AEON at Leipzig from summer 2026.
What is Toyota's automation philosophy?
Automate so that problems surface and stop the line, and keep people responsible for solving them. Quality comes first; labour saving follows from a stable process.
Company facts from Toyota, BMW Group and reports at the time. The reading of them is ours.
How useful was this article?
One tap. It tells us what to write more of.
About the author
SRF Capital Studio Research Desk
Funding Intelligence, SRF Capital Studio
The next one
Get what we publish next, by email.
Working notes on raising, borrowing, protecting, growing and structuring capital in India. One email a week at most, and you can leave any time.
We use your address only to send this. See our privacy policy.
We store your address to send you these emails and nothing else. See our privacy policy.
The services behind this article
What we do about it, for founders and finance teams. Each has its own page.
Process Standardisation and Optimisation
Making the way work happens repeatable, so it survives the people who invented it.
Debt & Blended Finance
Working capital, venture debt and blended structures for businesses that should not be raising equity.
