India is booming. The expanding population has overwhelmed the Bangalore-Mysore road the way a river floods its banks, and the flow of two-way traffic is choked with a living history of human transportation. There are belching herds of diesel trucks, diesel buses and iron-framed diesel tractors. There are wooden-wheeled carts pulled by brightly painted Brahma bulls, and two-stroke-motor rickshaws fueled by kerosene or cooking oil or whatever else is flammable and cheap. There are mopeds and bipeds and bicycles and motorcycles, and every conceivable type of petrol-powered, internally combusting automobile, from doddering Ambassador cabs to gleaming 16-valve Mercedes miracles. But there’s only one car like the one Somender Singh and I are riding in right now.

That’s because Singh invented it. Or rather, reinvented a piece of it: a small detail on the engine that he calls “direct drive.” He claims that his invention makes an engine
cleaner, quieter and colder than its internal-combustion cousins around the world–while using up to 20 percent less gas.

“Some people say to me, ‘Singh, why are you wasting your time on such a thing?'” he yells, his singsong Indian English barely piping above the tooting traffic. “But I tell you sir–I tell the world: I have conquered the internal combustion engine!”

To hear Singh tell it, his story has all the makings of a Bollywood movie, a classic heartwarmer about a small-fry Indian grease monkey who challenges the big boys armed only with a dream and a dirty wrench. And there’s no doubt that he has come up with something new, at least in the eyes of the U.S. Patent Office. But has a potbellied philosopher-
mechanic from Mysore really discovered the efficiency
El Dorado sought by every auto manufacturer, R&D center and thermal engineer from Detroit to Darmstadt?

Well, maybe. So far, all Singh’s invention has earned him is a few polite rejection letters from presidents, professors and auto manufacturers–while costing him tens of thousands of borrowed rupees and an untold number of sleepless nights. His eyes are glazed with the heat of an idea he can neither sell nor surrender. Mostly, he seems to have discovered the hard way that in 2004, it takes more than a patent and personal conviction to reinvent the automobile.

Even though Mysore is only a few hours south of the Indian IT epicenter of Bangalore, most of its 700,000 inhabitants lead traditional lives seemingly untouched by technology. The poor still work the fields and factories as they have for centuries, weaving silk or hand-rolling sandalwood incense; the last raja still lives in a whitewashed fairy-tale palace framed in stained glass and 97,000 lightbulbs. And every fall rich and poor alike make their pilgrimage up Chamundi Hill to pray to the mountain goddess who has watched over their tile-roofed city since time began. This is a place of yoga and vegetarian food, of barefoot men swathed in traditional white longhis and women draped elegantly in colorful saris.

In such a place, Somender Singh has long been an eccentric–a blue-jeans rock-‘n’-roller, a leather-jacketed motorcycle race champion and homegrown Evel Knievel, an
autodidactic birdman who soars above the palaces and red clay roofs in Mysore’s first and only motorized hang glider. Like most Indians, he is a reverent man; he prays to the mountain goddess for strength and wears a green ring from his guru to cool his fiery heart. But unlike most Indians, he also worships at the altar of the speed demon.

Singh has craved it for as long as he can remember: real bad-ass, teeth-gritting speed. And when, at age 10, a cricket ball to the eye destroyed his chances of following his father into the air force, Singh was destined to find that speed on the ground. So in 1968, following the time-honored tradition, he dropped engineering college for the old-school curriculum of trial and error and the dog-eared hot-rod canon of J.E.G.
Harwood’s Speed and How to Obtain It and Gordon Jennings’s Two-Stroke Tuner’s Handbook. He bought a motorcycle and then dedicated his health to racing it. When the sponsors wouldn’t touch him, he started his own team and called it Speedwell–and it did, winning more than 120 trophies for him as a racer in national and international events, and some 400 more for the machines he tuned.

Winning made Singh a local celebrity, to the point that when the first movie musicals were being made in the local Kannada language, the producers tapped him for typecast guest appearances–first as a pompadoured rocker strumming an electric guitar to Elvis’s “Hound Dog,” then as a daredevil motorcycle stuntman jumping stairs and cars. When a 1986 cycling accident left him with a broken shoulder and collarbone, Singh traded his helmet for a wrench and hung out his shingle as a mechanic. Now if you own a performance vehicle and are passing within a day’s drive of Mysore, Singh’s garage is a pilgrimage site of its own.

To get there, simply follow the Mysore road to a small sign announcing the headquarters of Speedwell Tune Up Centre and Garuda R&D. Beyond the sign you’ll find a little metal gate and a 50-foot yard containing a few cars, more motorcycles and the familiar open darkness of a working mechanic’s garage. Singh’s workshop doubles as the family home he shares with his wife and 10-year-old daughter. Out front, sleeping dogs and rusting car chassis lie in the shade of a rain tree. His assistants–four kids, their hair in modified d.a.’s, wearing rolled dungarees–peer into the mystery revealed beneath an open hood.

Singh’s office is in the back, separated from the greasy piles of engines and parts by a beige shower curtain. On the walls, competing for space between the pictures of Christian saints and Hindu gods and the standard mechanic’s warnings against urgent jobs and requested credit, are yellowed clippings celebrating Singh’s earlier life of speed. “Singh Takes the Day,” one reads, and “5,000cc Man Machine.” The grainy photocopies show a man who seems a lifetime younger, his eyes black and staring, his rugged mug framed with thick black hair. Below the photos, and a menacing poster of two jets about to collide in midair, is a sign bearing Singh’s motto: We specialize in work which few understand.
“And this has been my problem sir,” Singh says with a shrug. He settles in behind a metal desk heaped with paper and parts. “It has been my problem ever since I started this whole business of whatever I started doing in my life.”

It’s On this desk, somewhere Under the tools and parts and the notebooks crammed with letters and diagrams, that you’ll find a concave bit of steel, with rough grooves scored through the four axes like the points of a compass. It looks a bit like a homemade ashtray. In fact, it is Singh’s problem–his invention. Even as a prototype, it’s high-concept but exceptionally low-tech, the sort of thing you might be able to make in your own garage with a steady hand and a Dremel tool. Which is, essentially, what Singh did.

“I am no great genius man, no man with letters after his name or fancy institutions, and what I have invented is really very simple,” he admits, as he pushes aside the clutter to reveal a child’s chalkboard. “But to understand even so simple a concept, you first must have a basic understanding of the forces at work within the combustion cylinder, the concept of turbulence and combustion which define the engine.”

Singh takes the chalk and draws a rectangle with a domed top: a combustion chamber and the cylinder head, the ashtraylike piece of metal he has modified. Then he draws a diagonal line across the edge of that dome, then another, representing the grooves he has carved–his invention. The grooves are supposed to better mix the air and fuel inside the chamber. Singh is convinced that it makes combustion more efficient.

If a child’s chalkboard seems an overly basic tool for explaining a new engineering concept, remember that the internal combustion engine is itself hardly rocket science. Its fundamental conceit–a boom in a closed chamber, a zoom translated through piston, rod and crank–has remained pretty much unchanged since 1673, when the Dutch physicist
Christiaan Huygens designed a brilliant, nonfunctional, closed-cylinder, piston-driven engine that ran on gunpowder. The functional, liquid-fueled version of that invention–the internal combustion engine (ICE)–has been with us for about 200 years, over which time it has transformed itself from Swiss engineer Francois Isaac de Rivaz’s wonky four-wheeled hydrogen thingamabob (1807) to the fairly familiar gas-fuel innovations of Karl Benz and Gottlieb Daimler (late 1880s) and then wrapped under the familiar body shapes of Henry Ford (early 1900s).

Since then, gas has risen in octane, and the carburetor has been invented and largely discontinued. Engine emphasis has shifted from deep power to muscle to fuel economy and back, engineers have realized that compression is a key to maximizing thermal efficiency, and inventing the automobile has grown from an amateur’s obsession to a multinational juggernaut. But all of that is really just window dressing. The basic concept–the boom that turns a crank–has not really changed at all. And one of the physical fundamentals of that basic concept is turbulence.

Turbulence is the chaotic movement of fuel and air through the ICE’s combustion chamber–the swirl and tumble that makes hydrocarbons and oxygen combine fast and furiously in an efficient engine. Compressed fuel stagnates and separates and burns inefficiently, if at all–imagine
trying to burn a phone book without fanning the pages. Turbulence mixes it up, fans those pages. It’s what allows modern high-compression engines to go boom.

A hundred years ago, turbulence was to automotive engineers what chaos is to the Old Testament: a raw randomness ungoverned by words or math, an unordered whirlwind of particles as inexpressible to engineers as angels dancing on the head of a pin were uncountable to Sir Thomas More. Then came a Cambridge don named Harry Ricardo.
Like Singh, Sir Harry was a bit of an eccentric and was obsessed with motorcycles–though back in 1906, Ricardo was forced to create his own bike, by equipping his velocipede with a steam-powered engine fueled by coal fed from his own bulging pockets. As a proto?grease monkey, Ricardo intuitively recognized that air and fuel burn best when mixed. He then became the first to test the notion in the lab, measuring burn rates against the speed of a fan. The faster the fan, the better the burn; Ricardo had found the key to the boom.

Modern automotive engineers want turbulence, and they can describe it, just as modern mathematicians can describe chaos. What you want are swirling eddies of air and fuel mix, each variegated into smaller sub-eddies, and so on, down to individual molecules. Imagine it as a cascading Mandelbrot set of air and fuel inside the chamber. Then there’s a spark, and the whole thing goes off like a daisy chain of fire, a giant fractal fuse.

Engineers have devised all manner of technologies to create this particular form of chaos in their combustion chambers, from ornately angulated fuel injectors and domed cylinder heads to swirl-and-tumble-inducing atomizers. But 100 years ago, Ricardo found a far easier way to make the air-fuel mix in an ICE more turbulent. He built a combustion chamber that was domed in the middle and tapered on the edges, like a derby hat, so that the edges of the rising piston would come very close to the angled edges of the cylinder head. The piston goes up, and the fuel along the edges squirts into the center, to mix and swirl near the spark plug. Imagine pinching the edges of a jelly doughnut. He called this concept “quench.” Today we call it “squish.”

Squish! A laughably simple idea with a laughable name, but now almost every one of the billions of internal combustion engines operating around the planet employ some version of it–including virtually every engine Singh ever straddled in his 30 years in motorsports.

Singh knew that to get his precious speed he had to fire the heart of the engine, the center of its mystery: the combustion chamber. It was here that fuel was turned to bang–and here that the efficiency of that bang had stalled out at around 28 percent. The vast majority of the fuel was dissipated as engine heat or exhaust.

In the history of automobiles, manufacturers had experimented with all sorts of shapes and valve arrangements to improve efficiency, but nobody had ever dramatically altered the surface
of the chamber itself–perhaps, Singh reasoned, because engineers couldn’t see inside its metal walls and eyeball
its forces. The combustion chamber was a mystery
shrouded in plate steel. The very soul of the engine appeared ripe for improvement.

“From the beginning of time, whatever I did was geared toward taking an engine, polishing the rough edges out of it, and getting some more performance from it,” Singh remembers. “And I certainly knew that it was not God who was manufacturing these engines in a factory. It was just human beings, men set on a time frame, assembling parts. So there is, then, great room to improve.”

Singh needed his engines to work as efficiently as possible–he wanted the fuel to burn cleanly and under the
maximum compression. But like most tuners, he had run up against compression’s upper limit, above which pockets of unburned fuel explode spontaneously, or “knock,” under the pressure. He knew that the flame front from the spark plugs wasn’t reaching all the fuel at the edges of the cylinders.

One way to fight knock is with high-octane gasoline, which racers in countries like India have no access to. If Singh wanted more compression, he’d have to decipher the problem his way. So he started imagining: “My whole thing was, how on earth could one do something to mix it better?”

The simplest answer was Ricardo’s squish, which Singh, like many tuners before him, maximized into a sort of supersquish by making the rising piston head come as close as possible to the squish band. But the knock just got worse; either the chaos of the supersquish turbulence was too much, or the exploding hydrocarbons he was hearing were trapped inside the squish band, isolated from the spreading flame at the point farthest from the spark plug. The compression was stagnating his air-fuel mix. He needed to stir it up, to make that eddied, fractal fuse between the edge of the squish band and the center of the spark.

And so, armed with this intuition and a toolbox, Singh scratched his own small mark on Ricardo’s 100-year-old concept–through the squish band from the cylinder edge to the spark plug. Then he scratched another, and another. The first channels were shallow, and they quickly filled with hydrocarbons. Tentatively, he made them deeper. “We were very scared,” Singh confesses, and as he says it he sets down his nub of chalk in favor of a Gold Flake cigarette. “Maybe we were actually putting an induced crack into the head.”

But the engine didn’t crack. It changed. The compression went up, but the engine noise went down. And it seemed to be using less fuel: Measuring with a drip syringe and a stopwatch, Singh determined that it was between 10 and 20 percent less. “Most definitely and immediately, sir, something was very different,” he says. “My combustion was so stable that I could bring the idling down to such a point that you could actually count the blades on the fan as it turned.”

He felt the exhaust with his bare hand and noticed that it was running cooler. Yet when he removed the spark plug, he discovered that it had become blue, apparently from intense combustion-chamber heat. And when he ran his finger along the inside of the exhaust pipe, he noticed something else, or
a lack of it: unburned hydrocarbons. His engine seemed to
be running cleaner. In automotive terms, his squish-band channels seemed to have maximized combustion by propagating the laminar flame front from the spark plug to the edges of the cylinder at its top dead-center position, converting more fuel to expanding gases and piston work while avoiding the spontaneous combustion of unburned hydrocarbon emissions. In layman’s terms, they boomed better.

So much better, in fact, that he was able to keep his car in fourth gear at 500 rpms without sputtering or pinging, even while navigating the local congestion of bullock carts, rickshaws, bikes and cars. His engine ran so slow that it nearly didn’t need the gearing of a transmission–thus, “direct drive.”

He modified a motorcycle, then a two-stroke, then a four-stroke, then a car, then 50 cars. Finally he borrowed money from his mother-in-law and bought a spanking-new Tata
Indica in which to showcase his design. He decorated it with “direct drive” in stick-on letters on the steering wheel and a bull’s head above the grill. Then he tested his idea on a few customers, including N. Bhanutej, a writer for a national weekly newsmagazine who owns a pokey 1.2-liter Fiat Palio.

“Essentially, the whole car changed,” Bhanutej recounts. “It was zippier. And in third gear I could slow down to 20 kph with no engine knock, then press the petrol and just speed up smoothly, like you would in first gear.” He also found that his modified engine was strangely quiet. “At the stops, I sometimes needed to peek at the dashboard to make sure
it was still running. It seemed like a different car.” The mechanic at Bhanutej ‘s Fiat dealership thought so too. “He told me it was impossible for this type of car to perform this well,” Bhanutej says. “He kept asking about fuel additives.”

Singh seemed to be onto something. Although he couldn’t prove scientifically that it worked, he felt sure that it did. Certainly, it was novel–Singh applied for a patent in January 1999, and the U.S. Patent Office issued him No. 6237579 in May 2001. Two months after his application hit the patent office Web site, engineers from General Electric applied for a nearly identical patent for an aftermarket design, which they claimed, as Singh had, would result in increased turbulence, and thus better fuel efficiency, with fewer emissions.

“It’s very interesting, I think, that General Electric developed this idea after my patent became public,” Singh says with a smile. “But their design is very stupid. An add-on will never survive the intense forces of the combustion chamber. If I had come up with this idea, I would have been too embarrassed to tell anybody about it, let alone apply for the patent.”

This roadside mechanic in Mysore had seemingly beaten a billion-dollar R&D department. But what had he actually invented? Did it really work? Singh had his patent and his prototype. Now all that remained was to introduce his invention to the world.
So Singh wrote letters–dozens upon dozens of letters, each accompanied by an 8 x 10 glossy of his spark plugs. He wrote to presidents Clinton and Bush, to no effect. He wrote to Tony Blair and got a nice thank-you form on
10 Downing Street stationery. (“The British are a different lot,” he says proudly. “They respond to a letter.”) He wrote to
President A.P.J. Abdul Kalam of India and received a series of letters and promises for follow-up, none of which bore
fruit. He wrote to auto manufacturers from Dearborn to Pune, from Ford to Tata Motors. Tata, an Indian car company, expressed interest in the vague, noncommittal way that Singh had come to recognize as a mannered blow-off; Ford responded with a note wishing him “good luck,” which Singh didn’t much like, and the recommendation that he submit
his “suggestion” through the company’s dedicated Web site
(, which he liked even less.

Ford Global Technologies generates most ideas internally, employing 1,200 innovators–an alphabet soup of Bachelors and Masters and Ph.D.s from more than 60 countries, who file around 500 patents a year from gleaming Death Star?size facilities such as the Scientific Research Laboratory in Dearborn, Michigan, and the Forschungszentrum in Aachen, Germany. Outsiders like Singh are encouraged to submit through the Web site, and every year, 5,000 ideas pour in from inventors, academics, mechanics, customers and even children.

But of course, submitting with the masses was not Singh’s style. After all, he had conquered the internal combustion engine; he didn’t want to just click through a legal waiver and throw his life’s work, his lottery ticket, into a virtual wishing well, with no promise of return. Instead he wrote directly to the company president, and he did it by mail, with stamps and a typed letter and his standard spark plug photograph. He wanted to be recognized, singled out, and ushered through the front door. When he found himself repeatedly referred to the public portal, Singh simply took his business elsewhere.

Mostly, Singh spent his hope and energy writing to scientists. Surely, he thought, engineers would understand the significance of his idea! Or at least offer insight to what was happening inside his scratched cylinders. Singh writes the way he thinks, and his letters were excitable, florid documents in which his theories on combustion, turbulence and the environment were drawn in multiple colors and emphasized with triple interrobangs and exclamation marks.

The scientists’ replies were more compact. He claimed to have conquered the internal combustion engine? Using poor fuel on engines of antiquated design, evaluated without scientific instruments and in third-world conditions? Had he tested the design for 500,000 miles, they wondered, as a proper R&D lab would? He hadn’t–none of his
modified engines had done more than 65,000 road miles. Had he tested it on non-Indian vehicles or with the kinds of fuel used in the developed world? (He hadn’t.) Had he put it on a proper dynamometer, tested horsepower and torque? (No, but there’s a reason….) Could he send them an official printout from a five-gas analyzer indicating the oxides of nitrogen and carbon and the unburned hydrocarbons and total fuel economy? In a word, no.

It was possible that Singh’s invention was useful for the inefficient engines and poor-grade gasoline that crowd the Bangalore-Mysore road–but of course, any modern modification would improve on those ICE dinosaurs. So how, the scientists asked, did he know that his modification really did anything? Singh explained about the quiet and the low rpms, the blue spark plugs and clean tailpipes. “What more proof do we need?” he’d ask. “What more does the world need?”

As the scientists had made clear, what the world needed was proof of concept, in the form of hard, numerical data. But in Singh’s India, getting numbers is not as easy as you might imagine. First there’s the price: The most basic dyno test costs 25,000 rupees, or about $550, plus the cost of the engines, parts, assistants and fuel. That’s real money to amateurs anywhere; in India, where the average person earns around $250 a year, it’s real close to impossible.

Even if you can manage the money, testing in India is a difficult proposition. Singh repeatedly beseeched Mico-Bosch, a Bangalore subsidiary of the German dyno-testing giant, to let him pay for an afternoon’s test, and was just as repeatedly blown off. As he quickly learned, there are only three government-authorized dyno-testing facilities in all of India, each used almost exclusively for manufacturers. An amateur inventor here–even one with 25,000 rupees in his pocket–can’t just walk in off the street and test any old engine he likes, at least not without the written permission of the engine’s manufacturer.

“I imagined that these great men would say, ‘OK, let us get down to the bloody bottom line! Let us see about what on earth can be happening!'” Singh says. “Or perhaps, at the very least, be willing to take my money.”

The rule requiring manufacturer consent is apparently an effort to prevent individuals from disputing the official data on horsepower and emissions, as published by importers, manufacturers and the Indian government. “They don’t want any Ralph Naders popping up here,” Singh explains weakly.

In November 2002 Singh actually received one such permission from a manufacturer to test his modification on its engines. The manufacturer was Briggs and Stratton, and the engines were two 149cc side valves. Singh borrowed $3,000 and drove the 500 miles to the Automotive Research Association of India (ARAI) test facilities in Pune, but day after day, his test was delayed. He waited in a cheap hotel for two weeks, pacing, smoking, burning money. “It was a very frustrating experience,” Singh says, wringing the tension from his graying temples with permanently grease-stained fingers. “Sometimes it was like a bloody test of will.”

Finally he was allowed to bring his engines and hook them to a Benz EC-70 dynamometer with a five-gas analyzer and a Benz gravimetric fuel-measuring device. A week later, he
got his results. According to ARAI, at between 2,000 and 2,800 rpm, Singh’s modified engine used between 10 and 42 percent less fuel than its unmodified twin, with no appreciable losses in torque or power. And, as he suspected, it ran cooler too–as much as 16

An array of modified cylinder heads exhibit Singh's signature scratches--the grooves that, he says, enhance fuel turbulence and boast engine efficiency by upward of 20 percent.

Boom Time

An array of modified cylinder heads exhibit Singh’s signature scratches–the grooves that, he says, enhance fuel turbulence and boast engine efficiency by upward of 20 percent.
Singh's small office at the back of his garage, where he has composed dozens of letters trying to draw attention to his innovation. Left: The Tata Indica transformed by "direct drive." Right: The humble sign directing pilgrims to Singh's temple of automotive tuning.

Pilgrim’s Progress

Singh’s small office at the back of his garage, where he has composed dozens of letters trying to draw attention to his innovation. Left: The Tata Indica transformed by “direct drive.” Right: The humble sign directing pilgrims to Singh’s temple of automotive tuning.
Still life with Ambassador cab, down the road from Singh's shop.

Boom Time

Still life with Ambassador cab, down the road from Singh’s shop.