How to invest in the second wave of autonomous driving

The autonomous driving sector has seen an influx of investment in the past year after nearly a decade in the doldrums.

“We’ve been in this space about seven years now, and the last 12 months of activity dwarfs everything I’ve seen,” says Veerbhan Kheterpal, co-founder and CEO at Quadric, a startup developing processor architecture for the AI chips needed for self-driving cars.

Denso, the Japanese automotive components manufacturer, led a series B investment round in Quadric in 2022 and last October the two companies signed a license agreement, that will see them co-develop an in-vehicle semiconductor.

Quadric is by no means the only self-driving related startup receiving funding. Investment dollars in this area reached $9.06bn last year, nearly back to the level at the peak of the VC funding boom in 2021. 

Investments for the operators of self-driving fleets have been particularly eye-catching. Waymo, Google’s self-driving car subsidiary, raised a $5.6bn funding round in November, led by parent company Alphabet. It is offering commercial driverless taxi services in Phoenix, San Francisco, and Los Angeles with plans to expand to Austin, Texas, Miami, Florida and Tokyo.

Meanwhile, Wayve, a London-based self-driving car company, raised $1.05bn in a round involving SoftBank, Nvidia and Microsoft. Waabi, the autonomous trucking company backed by Volvo, raised $200m last June.

There have been plenty of casualties in the self-driving car market, including Argo.ai, backed by Ford and Volkswagen, which closed down in 2022. General Motors pulled the plug last year on Cruise, the robotaxi service it had acquired, although it still plans to use the underlying technology in other ways.

Special report:

Bumpy road: Self-driving cars make a cautious comeback

But investors say interest in autonomous driving is back, as technological developments make these systems a real, commercial proposition.

“It’s an exciting time right now in that all the work that’s been done, all the investment that’s been done over the last 10 years has finally coalesced into what I would consider the first wave of autonomous driving,” says Tony Cannestra, corporate ventures leader at Denso.

Kheterpal and Cannestra joined GCV’s recent webinar Is now the right time to invest in autonomous driving? together with Josh Berg, managing director at Magna Technology Investments, the CVC arm of the Canadian automotive components company, and Mitchell Yun, at Hyundai’s Mobis Ventures, to discuss how this next wave of self-driving development would unfold and where the best investment opportunities are.

Here are six things we learned from the discussion.

1. Self-driving technology has come a long way in the past few years

“Ten years ago, when the hype really started, all you had to do was put on your PowerPoint that you were a lidar company, and you know, you could raise money,” says Berg. “Now we’re starting to realise that lidar is really good in some applications but not good for others. You need to be able to look through fog or difficult boundary conditions, different weather.”

A whole raft of different sensor technologies have been developed to aid with some of this. And there have been big advances in computer vision system, says Kheterpal.

“If you compare it to 2019 and 2024, there’s been significant improvement in computer vision perception, particularly that runs on device.” This, says Kheterpal, has allowed “better prediction of pedestrian, vehicle behaviour around the car.”

Large language models, too, have started to be integrated with computer vision in the past 12 to 18 months, says Kheterpal. It is these developments that are allowing cars to start approaching the level four and level five stages of autonomy, in which the car can operate mostly or completely independently of humans.

In tandem, more advanced chip technologies like Quadric’s allow the heavy-duty processing that these types of systems require to happen in the car.  

2. The investment landscape is different now

The first frenzy of investors’ excitement over self-driving cars started in 2016 when GM acquired Cruise, says Berg.

“I was at GM at the time. That served as a catalyst for an explosion of investment into autonomy. That continued for a number of years because there was an exit [for self-driving car startups],” he says. Many of the large auto manufacturers acquired these startups, as well tech companies like Google and Amazon. Tier one suppliers to the automotive industry also acquired the technology because that’s what their car maker partners were looking for.

But now, says Hyundai’s Yun, “I think most of the exits are done. There’ll be some aqui-hires for smaller startups, [but] I don’t see a reason to buy the entire company.”

Hyundai has a majority share in autonomous driving company Motional, but, says Yun, automotive OEM’s like his are now putting so much R&D money into these partnerships and previously acquired companies, that they are not shopping for more businesses to buy.

An investor backing an automotive startup, therefore, has to be aware that there may not be any path to an exit and quick VC-multiple gains in the near future. They should only come in if they are interested in the strategic benefits of the technologies themselves, says Berg.  

“You have to ask yourself, why are you investing? There’s certainly money to be made, but you should be thinking about a product that you’re developing and investing in a company, because their technology can help you launch a real product,” he says.

3. Regulation is (still) the big bottleneck for self-driving

Regulation for self-driving cars is developing at varying speeds. In the US, certain states — chiefly California, Nevada and Arizona —  allow some degree of autonomous driving, but there’s not much movement at the federal level on regulation.  

Whether this would change soon is hard to say. No-one knows quite what US president Trump thinks about autonomous driving. He was not especially favourable to electric vehicles in his first term, but he may smile upon self-driving vehicles given his close association with Elon Musk, who has declared hopes for his Cybercab robotaxis to be on the streets by next year.

Cannestra says that the US Sun Belt might be a region where regulations on self-driving could start to be unified, allowing interstate driving for autonomous vehicles. “Texas, Arizona, I’m sure New Mexico would follow Nevada and California. They all seem to be thinking along the same lines of how to make more progress in these areas,” he says.

The EU is moving slowly but making progress towards unified automated driving regulations, notes Kheterpal.

The one place where regulation has moved fast, however, is China. “China has been particularly proactive, designating testing zones in most major cities and creating a clear framework for moving from testing to commercialisation,” says Yun.

4. China is setting the standard in all ways on automated driving

“We kind of woke up six months ago and realised we were behind in Europe, in North America, Japan, South Korea. All of the major automotive hubs globally really woke up to this notion that we’re all behind,” says Berg.  “With the governmental structure they [China] can think long term, and they can move quickly and not [face] nearly the same level of bureaucracy.  We’re seeing that play out in the market with incredible vehicles with very high tech and autonomy and electrification.”

Chinese leadership extends to every aspect of self-driving cars, adds Yun. That includes things like in-car experiences. “The technical experience inside of a Chinese vehicle is far and away better than what all of us here are driving on a daily basis,” he says. That does create an opportunity for investors, he says, because North American and European carmakers will need to catch up.  

 “Once the North American market and European market gets exposed to that, consumers are going to demand it,” he says.

5. New investment areas: embedded compute, thermal management, e-commerce

If there isn’t a quick buck to be made from flipping self-driving car startups to a corporate buyer, where are the best investor opportunities?

Quadric’s Kheterpal — perhaps unsurprisingly — sees advanced automotive chips as a key area to find undiscovered gems. “The embedded compute space is still wide open for a who’s going to emerge as the volume leader that actually ships in vehicles,” he says. “Even Nvidia, actually, is not at volume of edge devices. So that’s still an unsolved problem, where we’re seeing investors, deep tech investors, taking a look.”

Cannestra says there are also opportunities in solving the problem of thermal management in electric and autonomous vehicles. “If you combine autonomous vehicles with EV, the very nature of thermal systems has to change. You essentially need a complete redesign of the thermal unit in an EV, and that then leads to interesting design possibilities in the vehicle itself. So, there’s certainly opportunities there. More highly efficient electrical systems are going to be chased by a lot of tier ones and OEMs,” he says.

Yun says thermal management is an issue Hyundai is looking at. “EV batteries are hot, and heat is always a problem. So we’re looking for that space too.”

There may also be an another wave of interest in e-commerce in autonomous vehicles. “That was kind of a hot button topic in 2016/2017 but it never really took off. But, when you start getting level four, level five [autonomy], e-commerce in and out of the vehicle will be an interesting opportunity for OEMs to try and capture more revenue,” says Cannestra.

6. One of the offshoots of autonomous driving is robots

Those working on autonomous cars also see a strong adjacent market developing in robotics, such as in the hospitality sector.

“Lately we’ve got quite interested in human robotics,” says Cannestra. “I think there’s been an interesting push over the last two years to think about autonomy in all different kinds of sectors.”

Programming robots to do specific tasks in closed areas is much less complex than deploying automobiles to drive themselves at various speeds and conditions, he says. Companies that are cracking self-driving can apply the same technology to robots.

“We’re definitely seeing lots of similarities with robotics, both small delivery robots as well as humanoid robots,” says Kheterpal.  Humanoid robots also need complex AI models and high compute power. “Machines require low power, high performance, low latency to support AI models. In humanoid robotics we’re seeing very similar aspects to autonomous vehicles.”

Watch the full webinar below:

This webinar is part of GCV’s The Next Wave series of webinars. We run a webinar every month, alternating between advice for CVC practitioners and deep dives into specific investment areas. Our next webinar will be: Going Global – How CVCs can successfully expand into new markets Register here to secure your place.