Former Mobileye executive joins U.K. automated driving startup

Longtime Mobileye executive Erez Dagan joined Wayve, a U.K. startup with big ambitions in the automated driving realm, the company said Thursday.

Dagan will serve as president at Wayve, overseeing product, business and strategy for a startup that aspires to be the first to deploy vehicles using its self-driving systems in at least 100 cities.
This is a new position at the company. The role is similar to Dagan's post at Mobileye, where he was executive vice president of product and strategy before departing shortly after the global supplier's October 2022 initial public offering.

The similarities do not end there.

Like Mobileye, Wayve believes in an evolutionary path for automation, starting with driver-assistance features and building toward full autonomy. Also like Mobileye, Wayve believes in using cameras as the primary sensor in that effort.

Like Mobileye, Wayve collects data via cameras installed on cars and trucks already on the road to train its machine learning platform. It is now Dagan's job to accelerate pacts with automakers and other transportation companies that can make that happen at scale.

But there are substantive differences. Wayve has developed an end-to-end artificial intelligence approach to automated driving. With recent and rapid AI advances, Dagan said the startup is well-positioned to surge forward.

"It's a pivotal moment in terms of capacity, of generalization, of taking raw, unlabeled, unsupervised data and building a rich and actionable set of concepts to operate by," he told Automotive News.

That approach contrasts with the supervised learning path Dagan worked on in a variety of positions throughout his 20 years at Mobileye — in which humans label data and set up operational rules for a vehicle.

Generalized machine learning innovation now represents the best opportunity to solve the edge-case scenarios in autonomous driving that have, to date, stumped industry incumbents, Dagan said.

In a safety-critical application like transportation, of course, it is imperative that AI, unshackled from rules outlined by humans, make safe decisions at all times. Exploration into how to assure safe results, both at the U.S. Department of Transportation and by U.K. researchers, is ongoing.

Founded in 2017, Wayve has studied such an approach. Wayve executives believe it allows them to deploy vehicles in new geographies without maps. That's something of a polarizing topic in the industry, with Waymo, Aurora and others all relying on detailed mapping.

Wayve believes an unsupervised approach allows it to avoid the laborious labeling of training data.

"We're talking about taming the tiger," Dagan said. "You're taking the raw technology and making it an automotive product, and it's a fascinating journey."

Dagan sees potential even in the driver-assistance realm. The technology could allow for an automated driver assistant that mimics the driving style of a particular car owner. Or it could enable a human to have a conversational driver-assistance feature where a motorist tells the car to take an upcoming turn or stop at a nearby restaurant.

"When you can distill driving style from driving structure, you get unique selling points," Dagan said.

Wayve has offices in London and Mountain View, Calif. The company has raised $258 million, and is backed by Eclipse Ventures, D1 Capital Partners, Virgin, Baillie Gifford, Moore Strategic Ventures and others.