Another startup enters autonomous truck market with plans for retrofit system

user-gravatar Headshot
Updated Feb 27, 2017
Embark says its autonomous retrofit system should cost less than $50,000 to install on an existing truck.Embark says its autonomous retrofit system should cost less than $50,000 to install on an existing truck.

A new San Mateo, Calif.,-based upstart company has entered the autonomous truck ring with plans to develop an autonomous retrofit system for existing trucks that company co-founder Alex Rodrigues says should cost less than $50,000 per unit to deploy.

Embark, a developer of self-driving technologies for commercial transportation, Friday unveiled its self-driving truck technology, which uses a combination of retrofit radars, cameras and LiDAR depth sensors.

Embark is the second company in a year to announce plans to develop an autonomous retrofit system. Otto, now owned by ridesharing giant Uber, has run two on-highway tests with Volvo tractors equipped with its retrofit system, including a 120-mile trek in Colorado with a load of Budweiser in tow.

Rodrigues, Embark’s CEO and co-founder, said his company’s singular focus will make its product more trucker-friendly. “Otto is now actually Uber, which has a much broader focus and very different priorities,” he says. “Unlike Uber, we are 100 percent committed to trucking, so there’s no danger that our focus and our talent will get shifted over to passenger cars.”

Embark has completed more than 10,000 miles of testing with its fleet of five trucks and received its test license in the State of Nevada in late January.

Rodrigues says Embark’s truck is built specifically to handle long and simple stretches of highway driving. At the city limit, Embark’s computerized truck gives way to a human driver who navigates the city streets to the destination.

Embark’s technology allows trucks to drive from Interstate entrance to exit without any human input, he says. The data points from forward-mounted sensors are processed using Deep Neural Nets (DNN), a form of Artificial Intelligence that allows the truck to learn from its own experience.

By analyzing terabytes of real-world data, Rodrigues says Embark’s DNNs have learned how to “see through” glare, fog and darkness.

“We’ve programmed them with a set of rules to help safely navigate most situations, how to safely learn from the unexpected and how to apply that experience to new situations going forward,” he says.

Embark, according to Rodrigues, is working to perfect the underlying software platform that will make trucks smarter and more adaptable to new circumstances, and that can be easily integrated into existing or new – if even yet undeveloped – smart vehicle systems.

“We are committed to proving beyond a shadow of a doubt that this technology is safe and reliable,” says Rodrigues. “That means performing extensive tests and working with our partners in the government to get it—and the market—ready.”

As for when the market will be ready, Rodrigues says pinpointing that date is still a work in progress. The company is also still  determining a price point, but Rodrigues says he expects, at roll-out, it will cost less than $50,000 to outfit an existing truck with the company’s system.