Machines Made in India

The idea

Rohan is 26 and runs the night shift at a gearbox plant near Pune. Picture it a few years from now: the shop floor is dark and quiet. Robots built in Coimbatore load metal into machining centres built in Bengaluru, and every machine carries a quality mark that a buyer in Stuttgart trusts. Rohan runs twenty machines from one screen and earns far more than his father did on a single lathe.

Across the country in Dholera, a technician who grew up in a nearby village walks into a clean room where Indian wafers are made. Her chips leave for a railway signalling plant, a power-grid control centre and a phone factory in Noida.

In the Odisha steel belt, Kavita runs a materials startup. She walks ten minutes to meet the official who buys for her sector, and he can sign up to ₹500 crore on the spot. The test lab she needs is across the road, booked online at a published hourly rate. Industries grow where the buyer, the regulator, the test range and the talent sit side by side.

I believe a country that makes its own machines shapes its own future. Every factory on earth rests on three things: the machine tool that cuts the metal, the robot that moves it and the chip that controls both. Germany, Japan, Korea and Taiwan grew rich by mastering them. We have the engineers, the market and, with this book, the best place on earth to keep a profit, because there is no corporate income tax at all (Chapter 1). Now we can make the machines ourselves.

Why now

The prize sits in our import bill. India imported about $116 billion of electronics in FY26, up 18% in one year and about 15% of everything we buy from abroad. Every tenth of that import bill made at home is about ₹1.1 lakh crore a year of Indian output (estimate).

The machines that make machines tell the same story. India bought about ₹31,800 crore of machine tools in FY25 and imported about ₹18,700 crore of them, well over half. Each one built here keeps the skill, the service and the next order at home.

Robots show the gap and the runway. Korea runs about 1,220 industrial robots for every 10,000 factory workers, the highest in the world. India installed a record 9,120 robots in 2024, the sixth most in the world, yet our whole stock of about 52,570 is small beside the world's largest workforce.

The strategic sectors are ready for the same push. Our two defence corridors drew commitments of about ₹74,756 crore, and about 15% is on the ground so far. Lifting that to 40% adds about ₹18,700 crore of real investment (estimate), and the space economy alone is projected to reach $44 billion by 2033.

And we have already started on chips. The India Semiconductor Mission's ₹76,000 crore programme has approved India's first commercial fab, the ₹91,000 crore Tata and PSMC plant at Dholera, and Micron's assembly and test plant at Sanand began production in February 2026. Now we turn a start into an industry.

The big moves

Here is the path I see, one step after another.

1. First, make India the best place on earth to build a machine. There is no corporate income tax, so every rupee of profit can go back into the next plant (Chapter 1). Every imported part pays the same 5% duty, and a new plant self-registers and starts work, with strict liability and insurance in place of permits. That is the American lesson: freedom to build draws capital.

2. Next, let the state be the first buyer. Railways, defence, ISRO, atomic energy, power, public works and the builders who work for them buy certified Indian-made machines, robots and chips at a published fixed price, never above the latest import price, and pay within 30 days of inspection (Chapters 17 and 20). Each publishes a 5-year purchase plan that a maker in Rajkot can build a factory around.

3. Then set the bar at Europe's. One open specification, tested by any accredited lab, certifies Indian machines and robots against the safety rules of the EU Machinery Regulation, the law behind Europe's CE mark. As with Aadhaar, one published test lets in any maker who passes. A machine that passes in Pune is ready to sell in Munich.

4. Give every strategic sector a home city, chosen by a published score. We pick one lead city each for defence, space, biotech, new materials and nuclear energy, and a sixth for semiconductors and advanced machines. The weights are fixed before any state applies: supply base, logistics, talent, ready land and power, and security. An independent panel scores the bids, and every state's score sheet is public. Land goes by open auction under the Chapter 7 ready-plot rule and power at the one published tariff (Chapter 19), with ready land, firm power and recycled water at the gate, so states compete on readiness and we keep one country, one law. National funds co-invest in fabs, robot makers and machine-tool firms on commercial terms, alongside private money (Chapter 16).

5. Then move the buyers and the signing power to the builders. Within 12 months, each sector's operating bodies and the ministry wing that buys for it move to the sector city, led by an officer with full financial powers. A small policy cell stays in Delhi beside Parliament, and files move on a secure e-office. In each city, one existing government range or lab opens to private booking at published hourly rates within the year.

6. Finally, keep the keys at home and pay only for results. Chips for defence, the grid and telecom come from Indian-controlled trusted lines, and foreign partners are welcome as licensors and suppliers, as PSMC is for Tata today (Sovereignty first). Where the state needs a strategic capability, it buys it as a published outcome: for example a fixed fee for each trusted wafer delivered, open to any certified firm.

Today vs the new way

Here is how I see the new way comparing with today.

Together these turn India from a buyer of the world's machines into one of its makers, with six cities where buyers and builders grow side by side for about ₹3,300 crore of cash in year 1 (estimate).

Proof it works

Germany's family firms rule the machine shop. Germany's machine-tool makers, many of them mid-sized family firms of the Mittelstand, produced about €14.8 billion of machines in 2024 with about 65,300 people. Germany ranks second in the world in production and joint first in exports.

Taiwan built TSMC from a state lab. In 1987 TSMC was founded to turn research from the state-backed Industrial Technology Research Institute (ITRI) into a business. The government's development fund put up about 48% of the start-up capital and Philips about 27.5%, in return for its technology. A contract foundry that makes chips for others now holds about 70% of that world market.

Korea automated and grew richer. Korea runs about 1,220 robots per 10,000 workers against a world average of about 132, and remains a manufacturing powerhouse whose firms build robots as well as use them.

America bought back its chip industry. The US CHIPS and Science Act of 2022 set aside $52.7 billion, $39 billion of it for chip plants, and added a 25% tax credit on fab equipment. India offers a stronger pull than any tax credit: a profit here carries no income tax at all.

Bengaluru pulled in the space builders. ISRO's headquarters and centres drew much of India's 440 space startups to one city. The buyer's presence pulled the suppliers in.

Korea showed what to carry along. When Korea moved ministries to Sejong, officials kept commuting to Seoul because the signing power stayed there. Our sector cities learn from that and move the signing power with the people.

What success looks like

• In the first 30 days: the open machine and robot specification, mapped to the EU Machinery Regulation, the public buyers' 5-year plans, the sector city score and the trusted-chip rule are all published, and the five strategic sector cities are announced with every state's full score sheet.
• Within 1 year: at least three accredited labs certify machines and robots, and the first certified Indian machines are bought by public buyers at published prices.
• Within 1 year: the national funds make their first commercial co-investments in a fab, a robot maker and a machine-tool firm, and the Chapter 14 scorecard reports the share of public machine purchases made in India.
• Within 1 year: the moving units work fully from their sector cities by month 12, with one existing test facility per city open to private booking.
• After year 1: about 1 lakh direct jobs build up in the sector cities over 3 years (estimate), India's first commercial fab produces wafers, the import share of machine tools and chips falls every year, and Indian robots and machines sell into Europe.

Guardrail ideas

Ideas to start the conversation, for the builders, experts and the public to shape together.

• Every rupee buys a result. The state pays only for a certified machine delivered, a trusted wafer produced or a fair commercial stake, and every payment is published by firm within 7 days.
• Indian means made here. A machine counts as Indian when its maker publishes its local value added, checked by a randomly assigned auditor, and a false claim means the loss of certification and full repayment.
• The quality bar is Europe's, and it holds. Every certified model is tested against the EU Machinery Regulation, 2% of units are re-tested at random each year, and the maker carries strict liability and insurance for any harm.
• The keys stay at home. Chips and control systems for defence, the grid and telecom come only from Indian-controlled makers, and foreign partners join as licensors and suppliers.
• Every firm gets the same deal. Prices, fees and co-investment terms are the same for every certified firm, start-up or giant, with no cap on size.
• The choice of city is fair and visible. The score formula is fixed before bids open, and any choice other than the top scorer comes with written reasons laid before Parliament within 7 days.
• Critical capacity is spread for safety. One site holds at most 40% of any critical defence or nuclear line, with a second node at least 500 km away, picked by the same formula.
• Decisions stay in the sector city. The CAG checks every quarter that files are decided locally, with a published explanation if more than 10% go back to Delhi.

FAQs

Q: Isn't automation bad for jobs?

A: Automation grows the jobs that pay well. Korea has the most robots per worker in the world and remains a manufacturing powerhouse, because automated plants win the orders that keep factories open. Every robot also needs people to design, build and service it, and those jobs come to India when we build the robots here. Workers who change roles retrain at any institution with published job results (Chapter 15), and the one income-tested cash transfer supports any family in the change (Chapter 23).

Q: Won't chips and machines cost the taxpayer a fortune?

A: The taxpayer gains, because the state pays only for what it gets. Public buyers already spend on machines and chips; they simply buy certified Indian ones at a price never above the import price. National funds invest on commercial terms and sell their stakes later, and the trusted wafer fee is paid only for wafers delivered. The biggest pull costs the budget nothing at all: no tax on profit.

Q: Will the choice of sector cities be political?

A: The choice is open for you, and every citizen, to check. The formula is fixed before bids, an independent panel scores, every score sheet is public, and any change needs written reasons in Parliament. The second node in every sector is chosen by the same formula from other states, so each state sees exactly which scores to improve.

Q: Will officials end up commuting back to Delhi, as in Sejong?

A: Decisions get made where the work is. The buying and operating wings move with ₹500 crore signing powers, policy cells and ministers stay beside Parliament, and files move on a secure e-office.

The detailed blueprint for this chapter, a starting draft for the mission heads, industry and the private sector to refine, is in the Implementation Annex.