India makes zero of them.

Several years ago, engineers at a major Indian original equipment manufacturer (OEM) wanted to add exhaust gas recirculation to a production model. EGR is not exotic technology - the principle dates to the 1970s, and the hardware is a single valve, a position sensor, and a motor. A graduate student can design the control system in a couple of weeks. With model-based tools, it can be done in a day.

The engineers knew how to do it. What they did not have was access to the code running inside their own vehicle's engine control unit. The ECU - the electronic brain that governs fuel injection, ignition timing, and emissions - was supplied by a foreign tier-1 supplier. Its software was locked. Its calibration maps were proprietary. The OEM could not add a valve, change a parameter, or run an experiment without permission.

Three years later, the problem remained unsolved. Not because the engineering was hard. Because the key to the vehicle's brain was in Germany, not India.

This is not a company-specific failing. It is the defining condition of India's automotive electronics industry. The country that produced over five million passenger vehicles in FY2024-25 does not design the electronic control unit for a single one of them.

The Indian OEM is more comfortable writing a cheque to Bosch in euros than taking a chance on a domestic supplier. Indian companies are asked to be the best in the world in quality, quote 50 per cent or more lower, and prove success in global markets before an Indian customer will even consider them.

Every other successful automotive economy - Germany, Japan, South Korea, Sweden - does the opposite. They accept initial problems, invest strategically in local suppliers, and build domestic capability for long-term benefit. India's insistence on day-one perfection from domestic suppliers, while accepting decades of dependency on foreign ones, is self-defeating.

An earlier article in Swarajya, The Engine India Never Built, traced how India's auto industry - worth $250 billion and the world's third-largest - never developed indigenous engine design capability across seven decades. That was the heart of the vehicle. This is the brain. And in a modern vehicle, the brain controls the heart and everything else.

One thing to note is that when this article refers to "the ECU India never built", it means the most complex, highest-value electronic controllers - engine management units, vehicle control units, battery management systems. A modern vehicle has 50 to 100 ECUs, and Indian companies do produce some of the simpler ones. The core engine and vehicle control units, however, remain entirely foreign. That is the subject of this piece.

The Computer In Your Car

A modern car is a distributed computer system that happens to have wheels.

Think of it simply. Every control system has three parts: sensors that measure the world (temperature, pressure, position, speed), a controller that reads the sensors and decides what to do, and actuators that carry out the command (injecting fuel, opening valves, applying brakes). In a car, the controller is the ECU.

A mid-range SUV has 50 to 100 of them. The engine ECU governs combustion, the transmission ECU manages gear shifts, and the ABS and ESP units control braking and stability. In an electric vehicle (EV), the vehicle control unit orchestrates the entire powertrain - the battery management system monitors every cell, and the motor controller turns the driver's throttle input into torque.

Now consider this. Two OEMs buy the same engine and the same chassis. One tunes its adaptive cruise control to feel smooth. The other makes it aggressive. That difference - what makes a JLR feel like a JLR and a BMW feel like a BMW - lives entirely in the ECU software. If you do not control your ECU, you cannot differentiate your product. You are selling someone else's engineering under your own badge.

Electronics and software account for 30 to 50 per cent of a modern vehicle's cost, depending on the segment. More for a premium EV, less for an entry hatchback, but rising everywhere. India's automotive electronics market is estimated at $11 to $13 billion. Nearly all of it goes to foreign companies.

Who Supplies The Brains

The dependency is total - it does not matter whether the badge says Tata, Mahindra, Maruti, Ashok Leyland, or Force Motors.

Tata Motors uses Bosch engine ECUs across its entire ICE range. The 1.2-litre Revotron petrol runs on a Bosch ME17/MED17, while the 2.0-litre Kryotec diesel - a Fiat Multijet at its core - runs on a Bosch EDC17 with Bosch common rail injection. ABS, ESP, and airbag ECUs come from Bosch, Continental, or ZF.

In its EVs, the vehicle control unit - the top-level brain of the electric powertrain - is also Bosch, and the motor controller uses foreign power modules. The sole bright spot is Tata AutoComp's battery management controller, shown at Auto Expo 2023 - genuine Indian-designed BMS hardware. But even there, the cell monitoring chips are NXP or Texas Instruments.

Mahindra's ICE vehicles run the same Bosch family. But its new electric vehicles tell the sharpest story. Autocar Professional published a full supplier list for the XEV 9e and BE 6 on the INGLO platform. Vehicle control unit: Bosch, Germany. Electric drive unit: Valeo, France. Battery cells: BYD, China. ADAS controller: Mobileye, Israel. Software middleware: Vector Informatik, Germany. Airbag module: Continental, Germany. Not one major ECU is Indian.

Maruti Suzuki has never designed an ECU. Every engine management system across every model is Bosch or Denso. This is the same company that has never designed an engine in four decades.

In commercial vehicles, the picture is a Bosch monoculture. Ashok Leyland runs Bosch engine management, Bosch aftertreatment, and WABCO/ZF electronic braking across its entire fleet. Eicher, Force Motors - identical. No competition. No Indian alternative.

The scale of the dominant player tells its own story. The Bosch Group in India generated consolidated revenue of Rs 37,912 crore (approximately $4.2 billion) in FY2023-24 across 14 companies and 17 manufacturing sites. Bosch Automotive Electronics India - the subsidiary that manufactures ECUs alone - posted revenue of Rs 4,050 crore in FY2024-25.

Between 2002 and 2022, Bosch held the most automotive-sector patents filed in India at 156, according to GlobalData. The next four - Suzuki, Toyota, BorgWarner, Honda - are all foreign. No Indian company appears in the top five.

The One Exception

There is one Indian company that designs and manufactures ECUs.

SEDEMAC Mechatronics, founded in 2007 by IIT Bombay engineers, makes electronic control units for two-wheelers, three-wheelers, and generators out of Pune.

Its core innovation is a sensorless integrated starter generator (ISG) ECU that eliminates physical position sensors through software algorithms. The company holds patents on the technology and manufactures entirely in-house.

SEDEMAC's products are in millions of TVS and Bajaj vehicles. It holds roughly 35 per cent of the domestic ISG ECU market for two- and three-wheelers and over 75 per cent of India's genset controller market. In March 2026, it went public in a Rs 1,087 crore IPO.

By the measure of "does an Indian company design ECUs", SEDEMAC is proof it can be done.

But... TVS Motor accounted for 80 per cent of its revenue in FY25. Its products stop at sub-300cc two-wheelers and generators. It has not broken into passenger cars. SEDEMAC has been in the domain for over 15 years. It has the IIT Bombay pedigree, the engineering capability, and it is based in Pune - same city as Tata Motors. No Tata vehicle runs on a SEDEMAC ECU.

Why not? The answer has two sides.

On the OEM side, no Indian automaker has gone to SEDEMAC and said: here is a paid three-year programme, benchmark against this Bosch ECU, and replace it. They will not fund a proof of concept. They will not accept risk from an Indian supplier that they would happily accept from AVL.

On SEDEMAC's side, the company went for a blue ocean - sensorless control in the two-wheeler segment where Bosch had no product. Smart for survival. But it meant never building the track record or capability for passenger cars.

And rather than offering an open ECU platform that would let OEMs develop their own IP, SEDEMAC adopted what amounts to a mini-Bosch model - keeping the IP to itself. A different approach might have changed the conversation entirely.

The full scorecard is bleak. Tata Motors ICE: six major ECU types, zero Indian. Tata EVs: seven types, one partially Indian. Mahindra ICE: six types, zero Indian. Mahindra EVs: eight types, zero Indian. Ashok Leyland: five types, zero Indian.

Only in two-wheelers, where TVS and Bajaj use SEDEMAC, does Indian IP appear - giving those OEMs 25 to 40 per cent domestic ECU content. In passenger cars and commercial vehicles, the number is zero.

How The Lock-in Works

When Bosch sells an engine management system to an Indian OEM, it does not sell a component. It sells a system.

The package includes the ECU, the fuel injectors, the common rail fuel system, the sensors, the on-board diagnostics software, and the calibration data that makes it all work together. Buy the system and you get a vehicle that meets BS-VI. Try to buy just the injectors and develop your own ECU? You face massive resistance.

Bosch's margin comes from the bundle, not from any single part. A company that sells only injectors is a commodity vendor. A company that sells the system is a gatekeeper.

The software inside the ECU is a black box. The Indian OEM gets a PDF describing how things work, but it cannot see the source code, cannot change algorithms, and cannot add a function without permission.

Follow The Money

The real lock-in, though, is financial. It works through three layers of cost.

First, non-recurring engineering (NRE). Every new programme - a new engine variant, a new emissions tier, a new vehicle platform - requires NRE from the tier-1 to make the ECU work with the OEM's specific hardware. Industry figures suggest this runs in the range of Rs 10 to 30 crore per programme.

Here is a critical nuance. OEMs know that platform changes trigger NRE costs, so in recent years they have insisted on no changes to the underlying platform - only minor modifications on top, to keep NRE bills down.

This lowers the immediate expense, but it also means that in the era of software-defined vehicles, over-the-air updates, green hydrogen, and autonomous driving, rapid iteration and innovation are essentially out of the OEM's reach. They have prioritised short-term cost control over long-term capability.

What is worse, OEMs continue to refuse to fund platform development costs, expecting tier-1s to bear the entire investment. If developing a new ECU platform costs Rs 10 crore, and the OEM pays nothing, a new Indian tier-1 must amortise that cost over multiple programmes and multiple years - making it virtually impossible for a startup to challenge incumbents.

While the government exhorts Indian startups to build deep tech, how are they going to do this if venture capital will not fund new ECU platform development? And why would venture capital ever fund this if Mahindra & Mahindra and Tata Motors will treat Indian solutions with suspicion?

Second, per-unit hardware cost. An engine ECU runs roughly $250 to $500 per unit at production volumes, depending on the variant. A prototyping ECU can cost $1,000 to $2,500. These numbers do not sound enormous on their own. But Indian OEMs do not get the volume discounts that a Volkswagen gets, because their production runs are smaller.

And when Bosch repurposes software already developed for a European client and sells it to an Indian OEM, the Indian company pays full price for work that has already been amortised. The estimated premium Indian OEMs pay over European peers for comparable Bosch products is in the range of 20 per cent.

Third - and this is the one that explains everything - change request costs. Every time an OEM wants to modify anything in the ECU software, it submits a change request to the tier-1, even for trivial changes. The kind of costs reported in the industry are in the range of $500,000 per change request.

Think about what that means. If every modification to your own vehicle's software costs you half a million dollars, you become institutionally allergic to innovation. You stop experimenting, stop iterating, stop pushing boundaries. You accept what the tier-1 gives you, because the cost of changing anything is ruinous.

This is also why tier-1s resist any form of knowledge transfer. The moment an OEM or a local service provider can handle software changes and calibration updates independently, the change request revenue disappears overnight. The entire business model depends on the customer never learning to do things for itself.

The Calibration Dependency

Then comes calibration - the process of tuning thousands of parameters inside the ECU software to meet performance, emissions, and driveability targets.

To make it concrete: calibration determines how much fuel is injected at 3,000 RPM on a cold morning in Ladakh versus a hot afternoon in Chennai. It determines how aggressively the turbo spools, how smoothly the gears shift, how the engine responds at altitude. Getting it right across every condition is a two-to-three-year effort.

This is where AVL re-enters the story. Just as AVL provides the first 80 per cent of engine development for Indian OEMs - as documented in The Engine India Never Built - AVL handles calibration of the Bosch ECU software. The work is done from Graz, Austria, using the same platform AVL has already calibrated for Volkswagen or another European client.

The Indian company pays again, often at a premium, for work already substantially done. No technology is transferred to India, no engineering centre is built in Pune or Chennai, and the knowledge stays in Graz.

The Bosch-AVL combination functions as a technology duopoly over the Indian auto industry's electronic systems. Bosch supplies the hardware and base software, AVL tunes it, and together they control the full stack. The Indian OEM contributes the vehicle packaging, the marketing, and the assembly.

Around 2008, Mahindra took an XUV500 with a Bosch EDC17 ECU and calibrated the engine themselves - engine team and vehicle team working together between Nashik and Chennai. This was a step in the right direction. However, Mahindra has not made the natural progression - as Hyundai has - to owning the function software, or even enabling a local tier-1 company to make ECUs.

AVL continues to provide support to Mahindra's calibration programmes in various ways, and the Indian giant often refuses to entertain offers of calibration support from Indian vendors.

The end effect is that very few Indian suppliers think calibration is an area of work worth pursuing at all. Perhaps if the knowledge gained through the XUV500 programme had been leveraged differently, we could have seen the emergence of an "Indian AVL" by 2026.

As recently as 2025, an Indian engineering company offered a major OEM a calibration analytics platform designed and built in India. The answer: all our calibration is done in Graz. We have no control over it.

The Prototyping Gap

In Europe and the United States, the same Bosch systems are sold, but OEMs there also buy prototyping systems - commercially available from Bosch itself - that let their engineers hook into the ECU code, modify algorithms, test alternatives, and specify changes back to Bosch.

BMW, Volkswagen, and GM insist on this level of access. They own the calibration, they set the targets, and Bosch implements.

There is no technical reason Bosch cannot supply components and let a customer develop its own ECU. Caterpillar and Cummins have demonstrated this for years. Cummins has maintained its CORE2 technology for over two decades, allowing it to target Bosch hardware with in-house engine control software. This lets Cummins rapidly innovate and iterate without worrying about a tier-1 change request.

Caterpillar operates similarly. Both use Bosch mostly as a hardware supplier while retaining the control software IP in-house. No Indian OEM has attempted anything comparable.

The prototyping tools that enable this level of OEM independence exist and they are in Bosch's catalogue. On paper, Indian OEMs can buy them. In practice, nobody in India seems to have them, and nobody is working with them. Anything that would result in the Indian OEM gaining knowledge about its own vehicle's software is, for practical purposes, discouraged.

Many years ago, the EU took Microsoft to court for bundling Internet Explorer with Windows. Microsoft paid massive fines. Bosch bundles its ECU with injectors, common rail, sensors, calibration, and diagnostics. Why this bundling has never attracted anti-trust scrutiny in India is a question worth asking.

What Korea Did Differently

In 1987, Hyundai established Kefico as a joint venture with Robert Bosch and Mitsubishi Electric. The name stood for Korea Electronic Fuel Injection Corporation. The purpose was explicit: domestically produce electronic fuel injection systems. From the start, Hyundai treated the joint venture as a stepping stone, not an end state.

Kefico had its own research centre by 1992, and by 1996 it was performing system application and calibration - the same work Indian OEMs outsource to AVL in Austria thirty years later.

In 2012, Bosch exited and Hyundai became the sole owner. Kefico's own corporate history calls it "a difficult time of internalising the technology that was once dependent on advanced overseas companies".

But they did it. By 2015, Kefico was mass-producing its own GDI engine control units, and today it makes engine ECUs, transmission ECUs, hybrid management systems, fuel cell vehicle ECUs, and EV management systems. It ranks among the top 100 global automotive parts companies.

When Hyundai launched the E-GMP platform for dedicated EVs in 2021, the vehicle control unit, battery management system, and motor controller were designed within the Hyundai ecosystem.

Now notice the sequence. Kefico took over calibration first. That gave its engineers deep functional knowledge of the system. Once you spend a year or two tuning software functions across every driving condition, you understand the system inside out.

That understanding makes it natural to design your own software functions. And once you own the software, you can build your own ECU hardware to run it on. Calibration leads to software leads to hardware leads to sovereignty. It is a ladder, and the first rung is calibration.

The argument here is not that India should shut out Bosch or Continental. It is that India should do what every successful automotive economy has done: use foreign partnerships as a bridge to domestic capability, not as a permanent crutch.

Hyundai did not ban Bosch - it formed a joint venture, learned from Bosch, and eventually bought Bosch out. Toyota did not reject foreign technology - it studied it, absorbed it, and built Denso to exceed it.

The principle is reciprocity: India should support its own suppliers with the same conviction that Germany supports Bosch, Korea supports Kefico, and Japan supports Denso. Genuine competition, not protection, is what makes domestic suppliers world-class.

India, over the same 28 years in which Hyundai went from dependency to sovereignty, has remained exactly where it started.

BS-VI: The Opportunity That Was wasted

In April 2016, the Indian government announced the leap from Bharat Stage IV to Bharat Stage VI, skipping BS-V entirely. The deadline was April 2020. Europeans had taken roughly eight years to make the equivalent jump from a higher technological base. India got three and a half.

This should have been the forcing function. The compression should have pushed OEMs to develop local calibration and ECU capability.

The opposite happened. OEMs went back to the same foreign suppliers - AVL adapted existing Euro 6 calibrations to Indian conditions, Bosch supplied updated hardware and software, and the transition was completed on time. A real achievement in execution, but the result was deeper dependency, not less.

Indian startups that had entered the space found themselves shut out. One consortium - engineers with decades of experience at global OEMs and tier-1s - approached a major Indian OEM with a technology transfer proposition. The OEM took detailed notes. Then, to all appearances, it wanted its internal team of three engineers to implement what had been described.

Three. Not 500. Not 50. Three.

A manager at another major OEM explained the dynamic plainly: "If I say it's coming from AVL, the management might be willing to take the risk. If I say it's coming from an Indian firm with no track record, I won't be able to take that risk on my shoulders."

Perfectly reasonable from a career standpoint. But when every middle manager in every OEM makes the same perfectly reasonable calculation, the aggregate result is that Indian companies never get a chance to build a track record. The risk aversion feeds itself.

And when an Indian supplier does approach an Indian OEM, the response is: have you met all the European standards? Have you sold in Europe? Unless you have already succeeded in Europe, we cannot trust you.

But European markets protect their own through the same type of non-tariff, non-transparent barriers. Swedish companies promote Swedish startups. German companies promote German startups. Fraunhofer institutes bridge research to commercialisation.

Bosch and Continental are members of a German-led automotive software alliance whose explicit aim is to create a sovereign European software ecosystem, specifically formed to counter the growing dominance of American and Asian tech firms in vehicle software.

Where is the comparable collaborative effort among Indian OEMs to counter US and EU dominance in controls?

The chance to build an Indian ECU ecosystem during the largest regulatory-driven technology upgrade in the industry's history was lost.

From Engines To EVs: The Dependency Migrates

The EV transition was supposed to be the clean break.

An EV powertrain is fundamentally simpler than an internal combustion engine - fewer moving parts, less software complexity, and the control algorithms for motor torque and battery management are well understood. If there was ever a moment for Indian OEMs to take control of their electronic platforms, this was it.

The early evidence says otherwise. Mahindra's INGLO platform: six foreign-supplied major ECUs, zero Indian. Tata's EVs: Bosch vehicle control, Chinese cells, foreign motor controllers. The pattern from the ICE era has reproduced itself. The suppliers changed - BYD and Valeo alongside Bosch and Continental - but the structure did not. Foreign design, Indian assembly.

But the EV transition is only the beginning. The automotive industry is heading into an era of software-defined vehicles, over-the-air updates, green hydrogen powertrains, and autonomous driving. Each of these demands rapid iteration and tight control over the electronic platform.

An OEM that cannot modify its own ECU software without a $500,000 change request cannot participate in any of this. The need for in-house electronic control capability is not just about the current generation of EVs. It is about every generation that follows.

Indian automotive software vendors appear to have gravitated away from core engine control and towards infotainment, HMI, and communication software, where acceptance is easier to come by. There are promising signs of success in those areas, though it is concerning that Indian vendors are often forced to quote more than 50 per cent lower than a foreign vendor for the same piece of work.

Not every country's OEMs make their own ECUs - many use domestic tier-1s, and GM designs its own while Hyundai uses Kefico. The minimum expectation is that you have a domestic supplier designing the core electronics. That is what India lacks.

The Tier-1 Question Nobody Asks

When India discusses automotive self-reliance, the conversation focuses on OEMs. Can Tata build a world-class car? Can Mahindra compete with BYD?

The more important question is one tier below: why does India not have an Indian Bosch?

Rane Group, TVS Group, Motherson Sumi - large, capable companies that make wiring harnesses, valves, sensors, and brake systems to high quality standards. But if the ECU is the brain, these companies make the muscle. Nobody makes the brain.

Toyota spent decades building Denso into a nearly $47 billion technology powerhouse that designs proprietary components whose IP it owns. Hyundai developed Kefico, Hyundai Mobis, and a network of suppliers into independent technology companies. They invested in their domestic supply chains because they understood long-term competitiveness required it.

Indian OEMs have not done this. No systematic programme to upgrade tier-1s from mechanical manufacturing to electronic design. No embedding of engineers in supplier facilities for capability building. When Toyota partnered with ACMA to train Indian suppliers, it was a foreign automaker building capability that Indian OEMs had not.

An Indian OEM that asks the government for tariff protection but refuses to invest in a domestic electronics supply chain is asking for protection of assembly, not innovation. The same company that demands import duty protection does not extend the same willingness to its own supply chain.

What Would It Take?

The technical barriers are low. This needs to be said clearly, because for too long the Indian auto industry has told itself that electronic controls are impossibly difficult. That only Germans have the technology. That decades of expertise cannot be replicated.

None of this is true.

The algorithms that govern engine control, motor control, and battery management are well documented. Many foundational patents are from the 1930s. Closed-loop control theory is taught in undergraduate courses. What is so difficult about a PID controller?

Indian engineers at Bosch, Continental, and Denso design and calibrate ECUs every day, all over the world. The capability exists. It works for everyone except India.

The barrier is not talent, it is investment. India spends 0.64 per cent of GDP on R&D while South Korea spends 4.9 per cent, and that gap shows up in every metric. Indian auto component suppliers invest roughly one per cent of turnover in R&D; global suppliers invest eight to 10 per cent.

The path is not mysterious. It is the path Hyundai walked from 1987 to 2015, for instance.

Take control of calibration first. That is step one. It requires no technology transfer, no foreign permission. Just the decision to invest and to insist on access to the prototyping tools that already exist in Bosch's catalogue. Then take on smaller control subsystems - EGR, boost control, idle speed. Then the engine control unit itself. Then the full electronic platform.

Use the EV transition as the entry point. The control problems are simpler. The accumulated foreign advantage is smaller. This is where new entrants have the best chance. But the window is narrowing. China built its EV electronics dominance over 20 years of sustained investment. India's Production Linked Incentive (PLI) scheme is a fraction of that scale.

Indian OEMs should be encouraging their own engineers to start companies in this space. In any healthy ecosystem, they would be nurturing local suppliers rather than treating them with suspicion. If they had started 10 years ago, they would have better leverage with Bosch today. The supplier is more powerful than the customer. That is not sustainable.

Without an indigenous ECU industry, every adjacent ambition - domestic semiconductor fabs, software-defined vehicles, autonomous driving - lacks an anchor. A chip fab without a domestic ECU customer is a contract manufacturer. A software-defined vehicle strategy without control of the software is a branding exercise. The ECU is the middle layer that gives the chip a purpose and the vehicle an identity. It is the piece India is missing.

The 'Brain' India Never built

The history of Indian automotive electronics is a history of choices not made.

The choice not to acquire prototyping systems that would have given engineers visibility into ECU code. The choice not to fund domestic calibration when AVL was on call from Austria.

The choice not to develop tier-1s into electronic design houses. The choice not to use BS-VI as a forcing function. The choice not to use the EV transition as a second chance.

Each was individually rational. Buying from Bosch is faster. AVL's calibrations are cheaper than building your own. The foreign system works. The product ships on time. And if it goes wrong, nobody gets fired for choosing the German option.

But the cumulative effect of three decades of rational choices is an industry that cannot add a valve to its own engine. An industry where somebody has taken the key away. An industry that asks for tariff protection while sending its technology payments to Stuttgart, Graz, Tel Aviv, and Shenzhen.

South Korea started from the same place. In 1987, Hyundai needed Bosch and Mitsubishi to make a fuel injection ECU. By 2015, it did not. Twenty-eight years.

India has been manufacturing cars for seventy years. It has not started.

The engine India never built was a failure of ambition. The electronic controls India never built is the same failure, compounded. Because in the 21st century, the engine is just hardware. The controls are what make it think.

(With editorial contribution from Karan Kamble)