Why do we need this??????

These Billionaires Plan To Bring Self-Driving Tech To Everything That Moves

By Iain Martin, Alan Ohnsman

On a sunny winter day in Silicon Valley, Qasar Younis slips on his shoes in the entryway to Applied Intuition’s headquarters in Sunnyvale, California. The CEO’s strict ban on outside footwear in the building means employees and guests must wear slippers, a habit Younis picked up from a short stint at auto parts maker Bosch in Japan.

Feet now clad in black boots, he heads past a 150-foot water tower topped with what looks like a giant can of fruit cocktail—the last visible sign of the site’s roots as a Libby’s fruit cannery. His destination: the company’s garage, a tech warehouse in which you’ll find many of Applied’s latest experiments in car software.

Inside, a dozen or so young engineers who joined from companies like Honda, General Motors and Daimler line up next to various vehicles. Younis points to Jeep’s 2021 Grand Wagoneer, which is kitted out with its infotainment system, seating controls and vehicle diagnostics. Further down, there’s a robotic Isuzu box truck, currently being tested on Japanese highways. Nearby sits a small JCB Teleskid loader that can navigate construction sites on its own, and an autonomous Ford Raptor pickup, put together for the U.S. Army to haul gear without a soldier at the wheel.

Despite their differences, all these vehicles share software: Applied Intuition’s operating system, which Younis says can be used in any type of vehicle to connect and manage all its individual electronics and, increasingly, to drive itself. Founded in 2017, Applied sells the system mainly to traditional automakers like Stellantis, the successor company to Fiat and Chrysler, which inked a major deal in October. The pitch: Its tech will help them challenge next-gen players like Tesla, Google and Rivian, along with an emerging army of Chinese competitors that are turning cars into computers on wheels.

“Historically, [carmakers] acquire all these modules from suppliers for the braking system, the seats, and each one has a little software on it,” says Younis, a fast-talking Pakistani immigrant with a shaved head. “This is why they’re not like Tesla. They have to staple together five to eight different operating systems on a single vehicle, and you can’t do a single update.” Applied’s software, he says, “is the missing link to make cars and trucks and tanks intelligent.”

A one-stop tech shop is appealing for car manufacturers that have traditionally struggled with even the most basic software. Volkswagen, for instance, threw billions of dollars and thousands of engineers at an all-out effort to rival Tesla, only for the tech to be scrapped, the CEO to abruptly resign and the company left scrambling to integrate outsourced code. Even Ford and ever-reliable Toyota recalled over 2 million vehicles collectively last year to fix software glitches.

Applied booked around $800 million in revenue last year, twice as much as in 2024, while continuing to command gross margins of at least 80%. Top-tier investors BlackRock, Andreessen Horowitz and Kleiner Perkins love that growth and Younis’ vision (he’s “the best AI CEO nobody knows,” investor and board member Marc Andreessen recently posted on X). Last June, investors plowed an additional $600 million into the firm (it has raised $1.1 billion in all), pumping its valuation up to $15 billion. That means that Younis and his cofounder, Peter Ludwig, the company’s CTO, are worth at least $1.5 billion each, based on Forbes estimates of their stakes. Applied has been cash flow positive for “basically the whole history of the company,” Younis says.

Not content with making cars work better, Applied now wants to make them work by themselves. It already provides driver assistance as part of its operating system for autos, but Younis’ big bet is a single self-driving platform for everything that rolls, floats or flies, whether an F-150 pickup or an F-16 fighter jet. Its contracts now span the U.S. military, where it has $60 million in public deals with the Army and Air Force, as well as manufacturers that specialize in trucking, mining and construction equipment, like Komatsu and Swedish big rig powerhouse Scania. Ludwig claims that fully autonomous self-driving cars are just 18 months away, although a test drive shows Applied’s tech is still a bit rough around the edges.

“We have invested a huge amount in helping AI see, think and act. Now we can port that technology to all these different things,” Ludwig says. “Whether we’re talking about a car, a defense drone, a humanoid robot or an agriculture machine for farming.” That should help with the price, too—a single flexible system would cost less than sector-specific ones. Currently, Applied charges for its software on a per-vehicle basis. The specifics vary by contract, but a big manufacturer like Stellantis might pay more than $100 per car. Software for industrial vehicles, like mining trucks, is lower volume but more lucrative per vehicle.

Despite its excellent industry traction, Applied faces a range of competitors, including the likes of car software giant QNX. Tesla already has millions of cars on the road with driver-assist features that Elon Musk is itching to upgrade to full autonomous driving capability. And while Tesla lags Waymo in the robotaxi race, Musk’s tech is far more battle-tested than Applied’s. For now, Younis’ only self-driving client is Japan’s Isuzu, and some of his long-term customers like VW and Nissan are working with rivals Mobileye and Wayve instead.

But even if Applied’s software is ready in the next couple years, the combination of regulatory resistance and consumer caution means that self-driving cars on a mass scale are still far off. Instead, Younis is betting that bringing its autonomous software to spaces with more room for error will better position him to be able to win on self-driving cars too. Think mining haulers, farm combines or drone patrol boats, none of which need to worry too much about hitting pedestrians. It’s “radical pragmatism,” he says.

Younis and Ludwig grew up a few miles apart in the Detroit suburbs, where a nearby General Motors factory loomed large over their lives. Younis and his family emigrated from a farming village in Punjab, Pakistan, when he was in elementary school and his father landed a job in another local factory stamping out hoods and fenders. Ludwig’s father and grandfather, both GM men, discussed radar-based cruise control at the dinner table. “I probably took apart every electromechanical thing in the household to really understand how it worked,” Ludwig says.

The two never met as kids. Younis took the advice of his uncle, another GM veteran, and studied engineering at the “West Point of automotive”— Flint, Michigan’s General Motors Institute of Technology (now called Kettering University). He spent half his time learning about business, and the rest managing the V6 engine line. After his stint with Bosch car parts in Japan, Younis launched his first startup in 2007, a consumer crowdfunding business called Cameesa. It was a bust, but he struck gold with his second, TalkBin, a customer review app used by Crate & Barrel, that Google bought just five months after it launched (Younis netted a few million). He then joined the tech giant in April 2011 to work on Google Maps. A month later, he finally met Ludwig, who landed on his team after having studied engineering at the University of Michigan.

From Google HQ, the pair had a front row seat to the future of transportation. Just six months earlier, Google had unveiled the project that would become Waymo. “The natural extrapolation was, this is going to hit Detroit like a ton of bricks,” Younis says. “Like all good engineers in Silicon Valley, you think, ‘I should go do this myself.’ ”

Younis took his plan to Y Combinator’s Paul Graham. Instead of encouragement, he got a reality check. Robotaxis, starting a business and having his first child at the same time? Crazy. Instead, Graham convinced him to join the startup incubator in 2014, becoming chief operating officer, with Sam Altman (see “The Alchemist of AI,” page 60) as president. At Y Combinator Younis coached future unicorns DoorDash, Flexport, Gitlab and Cruise, the now-defunct self-driving car startup founded by Twitch’s Kyle Vogt.

When General Motors bought Cruise in 2016, Younis reached back out to Ludwig, who was still at Google. A significant shift was underway, he argued: “The car business is going to be a software business.” Google, Amazon and almost every carmaker in the world were now running self-driving experiments big and small. To capitalize on that, Younis and Ludwig decided to build simulation software— called “tooling,” it’s essentially a video game–like 3D world that can mimic millions of real-world situations—to test autonomous tech’s capabilities virtually.

They named their new outfit Applied Intuition and quickly landed deals with robotaxi startup Voyage (bought by GM’s Cruise in 2021) and self-driving truck company Kodiak, which now trades on the Nasdaq. “Tooling isn’t sexy, but it’s complicated and creates a lot of value right away,” says investor Bilal Zuberi, who led Lux Capital’s investment in Applied in 2017. Even today, it’s around a third of the company’s business, with major customers including GM and Toyota.

From there, Younis landed an early deal to build self-driving software for M1 Abrams tank maker General Dynamics. Now, later generations of the tech are being fitted to cars and trucks, with the operating system connecting everything from the brakes to seats to car computers. It acquired some of these capabilities by buying the intellectual property (and associated patents) of self-driving companies that went bust after running out of cash, such as autonomous truck developer Embark and OpenAI-backed Ghost Autonomy. “[Younis is] an execution machine like no other,” says Mamoon Hamid, managing partner at Kleiner Perkins, who first invested in 2019 and co-led last year’s $600 million round.

Today, some 1,300 people work at Applied on problems as mundane as Jeep’s lane guidance and as futuristic as U.S. Army robot trucks for combat. Younis doubled down on Applied’s military business with a 2025 acquisition of defense tech startup EpiSci, which helped make uncrewed F-16 fighter jets for the Air Force, and taught startup Aevex’s killer drones to fly in swarms. Applied’s tech garage even has a humanoid robot lurking in the corner (though Younis insists that’s not a big priority).

“They’re the most interesting company in autonomy,” says Grayson Brulte, founder of advisory firm the Road to Autonomy. “The fact that they can go across multiple verticals is huge. If one market goes into a recession, they’re not boxed in.”

On a bright December afternoon at an old gravel quarry in Sunol, California, about 25 miles from Applied’s office, a big construction excavator dumps a bucketful of dirt into the bed of a 40-ton yellow Komatsu truck. The truck drives hundreds of feet to a makeshift dump site, unloads and heads back to the excavator to do it all again, another dozen or so times before the day ends.

One day soon, there won’t be anyone behind the wheel. For now, there’s a safety driver, plus a team of four Applied engineers stationed in a small trailer nearby, glued to high-definition video and sensor feeds that monitor every movement in these simulated “LHD” cycles (load, haul, dump).

Automated trucks have been used at mining operations for many years in a limited manner, says Joe Forcash, who manages Applied’s automated industrials group and previously spent 25 years developing robotic vehicles for Caterpillar. But they aren’t perfect. “If anything gets in the way, vehicles don’t know how to go around it,” he says. Applied’s system promises to be able to pilot its machines around common obstacles such as boulders and Buick-sized potholes.

Building autonomous software for industrial trucks like this one is much easier than for vehicles that go on roads, where there might be random traffic cones, double-parked cop cars and texting pedestrians. Still, Isuzu, Japan’s commercial vehicle giant, is working with Applied on an autonomous delivery truck you might see heading crosstown as early as 2027, says Yasuhiro Yazawa, a general manager for Isuzu’s automated vehicle program. Initial testing of Applied’s system in Japan last fall showed promise—and problems. Merging and changing lanes still “need improvement,” he says.

Applied’s full self-driving tech for cars isn’t ready for prime time either, as a short test ride around Sunnyvale demonstrates. There are lags booting up the driving route and mild jitteriness similar to Tesla’s buggy “Full Self-Driving” system. Without public testing and scrutiny, there’s little outside technical analysis of just how good Applied’s platform is. At least for now, Stellantis, the world’s fourth-largest carmaker by volume, is holding off. Next year, Younis will earn a fee from every new Jeep or Peugeot rolling off Stellantis’ production line, but the deal to power the company’s “cabin intelligence” software doesn’t include any self-driving tech. “We will keep some autonomy development in-house,” says Stellantis CTO Ned Curic.

Company cheerleaders may think Applied’s all-in, autonomy-for-everything strategy is a good business hedge, but some critics say it’s a scattershot mess. “I think they’re in way over their heads, overpromising and chasing every latest shiny object,” says one industry veteran, who asked not to be named to speak freely.

If it does end up entering the self-driving car race, Applied faces a host of competitors beyond just Tesla and Waymo. Chip giant Nvidia announced in January that it’s building Tesla-like autonomous software, joining a field that includes companies such as Nuro, Mobileye and Motional. Another tough rival may be London-based Wayve, which similarly is developing all-purpose autonomous driving tech that can be used across a range of vehicles. “Our addressable market is every vehicle that moves,” says Wayve CEO Alex Kendall, who has raised $1.3 billion. Applied has lost one deal to Wayve already: Nissan, which uses its simulators to test driver assistance tech, decided to go with Wayve for its autonomy software. Younis says that Applied’s deal with the Japanese carmaker is “probably bigger.”

Thanks to its automotive operating system, its simulation business and a lead in automating “dull, dirty and dangerous” industrial vehicles, Applied has spread its bets wide. It’s a strategy that worked well in the last self-driving era, during which Applied thrived while many others failed. And there is no reason to suspect it won’t work again now, leaving Younis well-positioned to ride the autonomy wave, wherever it goes.