Google-backed Oxa drives alternative approach to autonomous vehicles

When it comes to self-driving cars, it is Google’s Waymo subsidiary —  which started offering commercial services in a handful of US cities last year — that tends to grab most of the headlines. But Google has another self-driving vehicle bet: UK’s Oxa, which develops software for self-driving vehicles. Google took part in Oxa’s $115m series C funding round in 2023 and has a 3.5% stake in the company.

Vehicles using the Oxa system have also started being deployed commercially in a number of locations in the past year, from retirement communities to airports and ports.

Paul Newman, founder of Oxa, says there is plenty of space for the startup’s alternative approach. “We’re not doing cars — we are a pure software play,” he says.

The company avoids the personal car and taxi market that many of the self-driving car companies like Waymo target. “We’re everything but taxis and personal transport. We’re buses, we’re mining, we’re industrial autonomy, ports, airports,” says Newman.

This niche approach, he says, allows Oxa to progress faster on issues like safety. The company plans to spread to more areas, much as Amazon did, which went from selling books online to broader e-commerce, web services and entertainment. The Amazon metaphor is apt as Gavin Jackson, who joined Oxa as chief executive in 2021, previously worked at Amazon Web Services.  

“The self-driving vehicle prize is gargantuan,” says Newman. “There is no doubt that there is an economic productivity, safety and business-building opportunity that comes if vehicles didn’t have to have one operator for every vehicle.”

Self-driving cars is an area that has recently seen a pick-up in interest, with investors putting some $9bn into autonomous driving-related startups last year, similar to the sums seen during the VC funding boom in 2021.

Oxa has been working on the problem of self-driving since 2014, when the business first spun out of the University of Oxford under the name Oxbotica. But now, as more self-driving cars have begun to appear on roads and in commercial deployments, it has grown easier to attract investor attention and raise money, says Newman.

Whereas even three years ago people still doubted if the technology would ever be viable, “now I don’t think anyone thinks that’s not technically going to get there,” he says. “There are things that remain hard, but no one is saying it will never happen. The conversation is more intelligent, it’s about where and when.”

Oxa’s other backers include Tencent, Ocado, ENEOS and BP Ventures. It is now working on raising a series D funding round.

Driving in a lot of “somewheres”

Unlike many rivals, Oxa has taken a frugal and minimal approach to how it builds its self-driving software.

“The worst thing you should do is have many vehicles driving around. There was this era when everyone was raising funding and companies in the States were spending billions on fleets of vehicles to get the data. That’s a terrible idea. What you want to do is build the technology that means you have to drive the fewest number of miles,” says Newman.  

Oxa hasn’t set out to create self-driving software that can be used anywhere. Instead, it focuses on becoming completely expert in specific locations. Eventually, says Newman, Oxa’s software will be used to “drive everywhere” by first driving in “a lot of ‘somewheres”.

“We’re very clear that we don’t solve the problem for everywhere. We don’t have any customers that are saying, ‘Please drive everywhere’. We’ve got customers who say: ‘I want to be great at package delivery around Heathrow’,” says Newman.

Oxa recently partnered with DHL to test autonomous driving at Heathrow airport, to see if the technology could be used for parcel delivery and baggage handling.

The company’s Oxa Driver service has so far been rolled out by US autonomous driving company Beep in senior communities in California and Florida. Oxa was also selected to provide the software for self-driving transit shuttles in Sunderland in the UK, and in Belfast harbour in Ireland.

Safety case

One of the areas where Newman says that Oxa can get ahead of competitors is safety. For self-driving vehicles to become mainstream, regulators will have to be convinced that they can drive without risk.

One of the most difficult pieces of this, says Newman, is showing that a vehicle can recognise when it is in an “unknown unsafe” situation.

“If you knew you’re unsafe, happy days. The sketchy thing is if you think you’re just nailing it and you’re not,” he says.  “I think that that’s being swept under the carpet by quite a lot of showbiz companies.”

Newman says Oxa builds recognition of “unknown unsafe” situations because it focuses on specific locations.  “I don’t have to say something about every traffic light on the planet, I only have to say something about the traffic control conditions at Heathrow. I can rehearse those 64 under a billion different conditions. If at one time they look like not one of those billion conditions, I might be unsafe.”

Working with strategic investors

Because of Oxa’s focus on software, it has worked with corporate partners from the early days of the company’s existence. It has collaborated with online grocery platform Ocado since 2017, first testing autonomous delivery vehicles in London. In 2021, Ocado bought a stake in the business.

Newman says having strategic investors on its board has allowed Oxa to learn faster from their experience. “The one thing that I really crave is someone telling me something that I had no way of finding out myself,” he says.

Working with Google has helped the company refine its cloud-based software.  “The sort of the technology boost that we’ve got by being next to a technology giant like that is fantastic,” Newman says.

Oxa’s software that “teaches” self-driving cars particular routes is based in the cloud, with individuals vehicles or fleets connecting with this periodically for upgrades.  “The majority of the technology is cloud based. That’s where Google came in. They’re interested in the cloud piece.”

Newman is keen to have more strategic investors on board for Oxa’s series D funding round. In particular, he seeks investors who understand that self-driving is a unique opportunity – but not a quick win. It will take time to scale up fleets of self-driving vehicles and revenues will build slowly.

“Many people rock up with a preconception of what your spend profile should be, what your ARR should be. It’s hard to break away from that formulaic thinking. But we’re talking about a generational business and a generational transition,” he says. “It’s unlikely to appear like other businesses do at this stage.”