What's in a Name?
On precision, patience, and the companies the market overlooks
A Carpenter Who Changed The World
In 1707, four British warships struck the rocks off the Scilly Isles. 2,000 men drowned in a single night. The admiral was court-martialled and shot.
The ships had not been attacked. There was no storm, no enemy, no mutiny. They just didn’t know where they were.
In the open seas, sailors could find the latitude (North/South) by determining the angle of the north star, Polaris and the height of the sun at noon. They had been calculating this proficiently for decades.
However, with no landmarks and no reliable way to determine east-west position (Longitude), navigators had been guessing for centuries. That night, the guess was wrong and turned fatal.
Unlike latitude, longitude had no reference point. Sailors relied on "dead reckoning" to find longitude, which involved estimating their speed using a log line and compass direction. Though functional, this method was highly inaccurate because it couldn’t account for hidden variables such as tides and wind currents, thus leading to fatal errors.
The British government launched the Longitude Act of 1714. It offered a massive £20,000 prize (around £4 million in 2026) to anyone who could solve the problem of determining a ship’s longitude at sea.
Time as a map: John Harrison, a humble carpenter and self-taught clockmaker from North Lincolnshire, realized that longitude could be calculated using time. The Earth rotates 360 degrees in 24 hours (or 15° per hour). If you knew the exact time at a fixed reference point, and you compared it to local noon (determined by the sun), the difference told you your position.
No such accurate clock existed back then. Not even a clock that could keep time to one second error rate per day at sea.
To put this in context, one second of clock error caused about 400 metres of position error at the equator. On a six-week Atlantic crossing, a clock that lost 1 sec/day would put you ~17 km off course - enough to hit rocks you thought were miles away.
Most clocks of that era failed at rough seas. Temperature changes expanded and contracted the metal parts. Humidity corroded them. The constant motion rolling and oscillation of the ship messed with the mechanism. Lubricating oil thickened in the cold and thinned in the heat. Salt air got into everything.
Harrison got his first breakthrough in 1727, as he built a clock made of wood that was accurate to one second a month. 50x more accurate than the nearest alternative. He called it Precision Pendulum No. 2.
Over the next 40 years, he built four marine chronometers - H1, H2, H3, and H4 - each one significantly better than its predecessor.
Harrison's H4, completed in 1759, didn’t even look like a clock. It looked like a large pocket watch, maybe five inches across. It was taken on an 81-day voyage to Jamaica in 1761. It kept time to within five seconds across the entire crossing.
To achieve this, he had to invent solutions in spring metallurgy, gear geometry, and temperature compensation. Note that these concepts barely existed in practical form at the time. He used a bimetallic strip to correct for temperature change by exploiting the different rates of expansion of brass and steel. He used lignum vitae, a hardwood from the tropical America. The wood is so oily that it was self-lubricating, thereby eliminating the need to oil the clock.
What Harrison proved is the thing that matters most - the accuracy of a mechanism is only as good as the accuracy of its individual components. You cannot achieve one-second-per-day timekeeping accuracy with parts made of five-second-per-day tolerances. The precision in the system requires precision in the parts. Precision propagates upward, or it doesn’t!
Two centuries later, in Stuttgart, an engineer faced a different problem.
His name was Robert Bosch. His company had built its fortune on magneto ignition systems. Ignition magneto is a dated yet profound ignition system that used electrical spark to fire petrol engines. By the early 1900s, Bosch magnetos were in cars, trucks, and motorcycles across Europe.
Before the diesel engine became practical for vehicles, moving things was expensive, slow, and limited by geography. Goods traveled by rail or horse. Farms were local because food couldn't travel far before spoiling. The truck didn't exist as an economic force because the engines available were either bulky steam engines or underpowered petrol engines. They couldn't move meaningful loads efficiently by road.
The diesel engine changed this forever. Rudolf Diesel’s engine was more efficient than any steam engine - it extracted more energy from the same amount of fuel. But in the 1890s it existed only in large, stationary applications such as power stations, ships and factory floors.
The challenge was injection. Diesel engines didn’t need a spark. They ignited fuel through compression heat alone. Precisely injecting diesel in the right amounts thousands of times per minute was unsolved for a moving vehicle-scale engine.
He needed to build a pump that was small enough to mount on a vehicle engine. The pump would have to be powerful enough to inject diesel fuel at pressures high enough to atomise it into a fine mist. It had to be precise enough to meter microscopic quantity of fuel in fractions of a second. If achieved, it would eliminate the bulky air compressor that made diesel engines impractical for vehicles. It would make the diesel car possible.
The heart of the pump was a plunger sliding inside a barrel (like a mechanical bicycle tire air pump). The gap between them had to be one to three microns (µm) . (One micron is one millionth of a metre. Average human hair is 70-100µm.)
Without using rubber or a gasket, he had to ensure that the gap was precise enough so that high-pressure fuel cannot pass through. At that tolerance, the fit itself becomes the seal.
Bosch’s engineers Franz Lang and Walter Lippart spent five years solving it.
The machine tools barely existed. But the real challenge was the measuring instruments required to verify that 1-3 µm gap — they didn’t exist. You cannot make something to a tolerance you cannot measure. So before producing the pump, they had to develop the gauges. Before they could build the gauges, they had to build the reference standards to calibrate the gauges against. They had to climb the ladder of precision from the bottom, building each rung before they could reach the next.
The grinding machines capable of achieving consistent cylindricity at production volumes did not exist either. Bosch developed them. Interestingly, each plunger was matched to its specific barrel. They are made in pairs and not interchangeable. This had to be codified from craft knowledge into an industrial process - one that could be taught, repeated and verified at scale.
It took them years. When they succeeded, they had not just built a diesel injection pump. They had proved that sub-micron precision was achievable in industrial production.
In 1927, Bosch approved the pump for large-scale series production. The first thousand units went to MAN in early 1928. By 1934, they had produced 100,000. Within a decade, nearly every European truck manufacturer had converted to the Bosch system. When the first diesel passenger car arrived in 1936 - the Mercedes 260D - it ran on a Bosch pump.
Both stories are the same story.
A product that demanded the level of precision that did not yet exist. The people who built it had to first build the capability to build it. The capability, once built, became the moat - because the next person who wanted to make the same thing had to climb the same ladder, and the knowledge accumulated in the climbing could not be bought or copied easily.
In this series of posts, I will share my thoughts about precision engineering companies in India that are overlooked or somewhat misunderstood.
Industrialising the Craft
When those first thousand pumps shipped to MAN, they didn’t just solve an engineering problem; they represented something rarer: the industrialization of a craft.
Before Bosch, achieving a 1-3µm fit between two moving parts was the domain of master watchmakers and scientific instrument builders. It was done by hand, one piece at a time, by people who had spent a lifetime learning to feel what a micron was. It could not be taught from a manual, let alone replicated at scale.
Bosch shattered that constraint. While apparently building a pump, the company developed the manufacturing system that made the pump replicable.
First, it solved the metallurgy. Standard steel wasn’t good enough. Under the pressures and temperatures inside a diesel injector, ordinary alloys would deform. Bosch developed specialized alloys that wouldn’t deform or seize, even in the near-zero gaps.
Second, it solved the geometry. Shape is equally important to size when dealing in microns. A plunger that is exactly the right diameter but slightly oval will bind; if it’s tapered, it will leak. Bosch perfected centerless grinding techniques that produced plungers that were perfect cylinders.
Third, it solved the calibration problem. In a multi-cylinder engine, every injector has to deliver exactly the same amount of fuel. A one-percent variation creates uneven combustion, vibration, and power loss. Bosch developed the test bench. The test bench was a device that measured fuel delivery from each pump element under controlled conditions. Every pump that left the factory had been proven on the bench, not just measured on the drawing.
This three-part system - material, geometry, calibration - is still the foundation of precision manufacturing. It’s what separates a company that can make one good part from a company that can make a million identical ones.
What Precision Engineering Actually is
Precision engineering is a specific capability, built through necessity, that compounds over time, and becomes progressively harder to replicate.
Precision engineering is not an industry or sector as many investors erroneously classify it. It’s a discipline of designing and manufacturing components with high accuracy (tight tolerance) and superior repeatability.
Every manufactured component has a tolerance. A dimension on a drawing has a permitted variation i.e. how far from perfect the part is allowed to be before it stops working.
The tolerance determines which machines you need and which processes you must control. That in turn will determine which customers will trust you and what you can charge them.
The unit of conversation is the micron. At that scale, even something which appears trivial to a layman - such as temperature - matters a lot. A 500mm steel part can grow 6µm for every 1°C increase in temperature. That 6µm difference in a critical application could generate enough heat to destroy the subsystem. That means a shop that doesn’t control its environment to within half a degree cannot hold a two-micron tolerance across shifts. The morning batch and the afternoon batch will be different!
Things an Investor Must Know
We don’t necessarily need to understand grinding tolerances to invest in this sector. We need five core ideas.
Precision exists on a spectrum.
At one end you get general manufacturing: thousands of companies compete on price here. At the higher end, you get fuel injectors, turbine blades, surgical implant surfaces - you can count the qualified global suppliers on two hands.
The tighter the tolerance requirement, the smaller the competitive field, the greater the pricing power, and the longer the customer stays. Every step up the precision ladder roughly halves the number of companies that can do the work and doubles the time it takes a customer to leave.
The international tolerance grade system (IT grades) provides a single number that describes how precisely a component's dimensions must be controlled relative to its size. Think of it as a credit rating for a manufacturing process.
Position on that spectrum determines your competitive field. Yet, compliance to IT grade doesn’t fully reflect capability. A company that produces IT4 components occasionally is a job shop with a good story. A company that produces IT4 components consistently, across batches, across shifts, across years is the kind of business I’m interested in. The icing on the cake is if this business is regularly verified by a stringent customer that audits every process.
Certifications are a proxy, not a verdict.
Certifications are the easiest filter. But you need to know what each one actually means.
ISO 9001 tells you the company has documented procedures. Thousands of Indian manufacturers hold it. Tells you almost nothing.
IATF 16949 is the automotive quality baseline. Indian manufacturers are increasingly chasing this. This standard separates the chaotic from the competent.
NADCAP is categorically different. It is a process-specific audit conducted by a retired aerospace engineer against the most demanding standards in global manufacturing. It is specific to a facility and a process and conducted every 12-18 months. It cannot be transferred and has to be earned, process by process, facility by facility. Some NADCAP certifications are more demanding than others. Measurement, Inspection, and Non-Destructive Testing (NDT) are the more accessible entry points. Coatings and Chemical Processing are harder. Heat treating and Hot Isostatic Pressing are even harder - few Indian facilities have this.
The customer relationship is the best quality certificate that exists.
You can’t visit every factory, but you can find a customer list and dig a little deeper. Not all OEM customers are equal. For instance, Bosch, BMW, Daimler, MTU are extremely demanding. Bosch wouldn’t stay with a supplier for ten years as a favour. Their engineering teams conduct process audits annually, score suppliers against hundreds of specific criteria, and disqualify those who slip. The German system uses a common framework called VDA 6.3. It’s a process audit so detailed that a score of 88% with BMW or Bosch is like entry into an IIT/IIM. Japanese OEMs - Toyota, Honda, Denso - have their own equally demanding standards, with proprietary customer-specific requirements layered on top of the global IATF 16949 baseline. Surviving these audits for over a decade means the company has been verified, repeatedly, by the most demanding quality systems in the world. Approvals such as these reflect more about quality standards than certifications only.
Equipment can be bought. Process, not so easily.
Any company with a few hundred crores can buy some expensive European machines, furnaces, and measurement instruments. The machines are a means to an end, not the end itself.
What cannot be purchased off the shelf is the knowledge of why those machines produce the results they do—the compensation model built from thousands of heat treatment cycles, the tooling library designed for a specific SKU, the material relationships that eliminate batch-to-batch variation. These usually take years. A company that owns only the equipment is likely a job shop.
What This Series Is and What It Is Not
Some context before we go further. The obvious names in Indian precision engineering are already discovered. Sona BLW, PTC Industries, Bharat Forge, Motherson Sumi, Tube Investments, MTAR Technologies, Data Patterns — these companies have heavy institutional ownership, and valuations that reflect the market’s enthusiasm for the theme. The discovery is by and large done here.
My clients and I invested in Sansera Engineering around ₹800, averaging up to approximately ₹1,200. The thesis was straightforward: a precision forging and machining business with a solid automotive base, growing aerospace revenue, and a management team that understood the difference between building capability and building a story. While the investment has done well, the present valuation has moved into territory I find difficult to justify with new capital. So Sansera will not appear in this series despite its growth prospects.
What I am looking for is businesses that the market has not yet looked at through the right lens. Companies that have built genuine process depth without boasting about it. Companies that may not have ‘Aerospace and Defence’ in their press releases or investor presentations, but have been supplying the most demanding customers in global manufacturing for years.
A word on aerospace specifically. While the theme is genuinely exciting, the reality is that aerospace programmes have long qualification cycles before the first purchase order arrives. Even after qualification, orders can be irregular. Programme ramp may take years. It affects revenues with very little ability to redirect capacity unless your plant and machinery are fungible.
My preference is for a manufacturer with a robust volume floor from automotive and niches in industrial + optionality of growth from Aerospace (same playbook as Sansera). The auto business provides volume and earnings predictability. A&D venture provides re-rating potential. If the A&D ramps up, it compounds. If it is delayed, the business is not impaired. Needless to say, it should have a fortress balance sheet and excellent working capital management. Heads I win, tails I don’t lose much. This is how I’m wired - I don’t prescribe it to others!
One more thing. The market’s enthusiasm for precision engineering has created a category of company that looks attractive but is vulnerable to shocks. Consider a generic drug manufacturer: it invests heavily in approved facilities will still compete on price with every other manufacturer that made the same investment. If the only barrier is capital/capex, what prevents competition from adding capacity, especially if the market is ready to pay premium for a narrative? High multiples without a defensible moat attract competition.
I am looking for entrenched customer relationship (demonstrated through market share), years of R&D commitment, preferably niches where no new entrant has been able to make a dent in the past decade. Process knowledge is the invisible moat that can’t be replicated by writing a cheque. You must endure years of grit and pain to build tacit knowledge. I prefer companies that have this kind of moat.
What’s there in the name?
I want to tell you what I kept finding as I went through companies claiming to practice precision engineering.
The ones with impressive decks are often disappointing on true capability. Some CNC job shops with limited process engineering knowledge flash images of fighter jets and tanks on their PR/IR material.
The ones that I liked conceal more about their business (trade secrets, processes, revenue mix, focus areas) than they reveal.
I'm grateful to Mr. Laxman Katakkar (CEO, Uni-Tech Automation) and Mr. M N Kumar (Fellow ValuePickr Investor with deep engineering knowledge) for conversations that shaped my thinking on this subject. Any errors in understanding are mine alone.
The next post in this series is about one such business. It doesn’t advertise itself as a precision engineering firm. It has a near monopoly in its niche while serving the most demanding customers in auto.
Great write up.
Hardly many of them were not aware what is precision Engineering means. Thanks for the detailed note.
So just to understand which are the companies that the market are not understood or un discovered players. thank you.