#173-Noam Arkind, Arbe Robotics

Marc Hoag AI Law Pod (formerly Autonomous Cars with Marc Hoag)

My guest today is Noam Arkind, the CTO at Israel-based Arbe Robotics (@Arbe_Robotics). Noam holds a PhD in Applied Math from the Robotics Lab at the Weizmann Institute of Science, so it’s no surprise that we did a fantastic one hour deep dive into all the fascinating technology behind their 4D high-resolution imaging radar solution.

What struck me most about Arbe — and it took me until maybe half-way into our discussion to realize it — isn’t just their full development kit that includes both the hardware and software; it wasn’t just the cost competitiveness over alternative solutions like lidar for instance; but rather, it was the recognition that Arbe is quite literally shaping up to be a better lidar than lidar, from virtually every metric.

Where computer vision cameras are limited by today’s AI and ML capabilities; where traditional radar is limited, at a minimum, by resolution; where lidar is limited by cost, inclement weather, and potentially resolution too, Arbe solves all those issues by effectively producing a 2K radar solution they claim is “100 times more detailed than any other radar on the market, suited for every level of autonomy or class.”

And that last point is what’s most exciting, I think, about Arbe: their solution isn’t designed for future L4 and L5 vehicles some five to 15 years hence; on the contrary, if only because the physical hardware is the same form factor as existing ADAS-equipped vehicles’ radar kit, Arbe can easily be used to augment or fully replace existing automakers’ ADAS solutions and get safer vehicles on the roads today.

Arbe has raised $55M to date, and is based in Tel Aviv with offices in Plano, Texas and Beijing.