Tesla CEO Elon Musk famously hates lidar, the radar-like laser-powered sensing technology for autonomous vehicles, including self-driving cars. At Tesla’s first-ever Autonomy Day in April 2019, Musk famously said that anyone relying on lidar is a “fool’s errand” and that anyone relying on it is “doomed.”
Not shockingly, the CEO of a lidar company disagrees with Musk.
But not entirely.
“Five years ago, when he had to make a decision on a car that he brings to the market, there was no lidar that was available at the right price point and performance and maturity,” InnovizCEO Omer Keilaf told me recently on the TechFirst podcast. “So he made a good decision.”
But now, Keilaf says, reality has changed.
Lidar components that used to be tens of thousands of dollars are now just $1,000, and there’s a clear path to getting them down to $500 soon, and $100 in the foreseeable future. Startups like the Israel-based Innoviz, which recently went public in a reverse merger via a SPAC that raised $350 million of capital, think that’s the future.
And customers like BMW, Keilaf argues, prove that lidar is the best and safest choice.
The key point for Keilaf?
Environmental hazards for camera-based systems like raindrops and mud, which block computer vision but do not impede, he says, Innoviz’ Lidar-based systems.
“Cameras can achieve quite well in seeing objects, as long as they see,” he told me. “And when they don’t, there is no other sensor that can reach that. Even the most advanced radars that are in discussion today are far below the needed resolution. So, it’s only LIDARs.”
The InnovizOne is a solid state lidar that mounts in the car grill. It can see objects from ten centimeters away to 250 meters, about 820 feet. With a maximum field of view of 115 degrees horizontally and angular resolution of just .1 degrees, it can distinguish small details. And at five to 20 frames per second it’s faster enough, Keilaf says, for cars traveling at high speed.
The resolution is key.
While visual systems for self-driving cars currently outperform the human eye, and radar-based systems like Occulii are now at sub-one-degree resolution — a massive improvement for radar — lidar advocates say that its range, resolution, and resistance to weather conditions make it the technology to beat. (Of course others argue that lidar is not as resistant to heavy fog, snow, dust, or mud as its advocates believe, meaning that self-driving might only be possible in fair weather conditions.)
Keilaf says lidar is the best technology because it provides better insight into what’s around your autonomous vehicle.
“I think a LIDAR has a very strong benefit which gives you a very good understanding of the scene for very different types of objects,” he told me. “A camera is mostly relying on the fact that you are able to train the system to identify what it is in order to interact with it. A LIDAR is actually a physical sensor. You send a pulse of light and you measure a reflection, a light reflection from an object. So you basically can see it. So you know where it is. You know what’s the size of it, even if you don’t exactly know what it is … I mean, if you didn’t train your camera to understand what is that object … not sure how would you react to it.”
One big player that hasn’t really shown its cards yet is Apple, and Apple is rumored to prefer lidar. (In fact, it’s apparently already in talkswith lidar manufacturers.) That’s good news for Keilaf and other lidar makers, but Apple is years away from even unveiling what its autonomous vehicle aspirations might look like.
Ultimately, all the technologies are getting better very quickly: visual, radar, and lidar. For Keilaf, that means multiple lidars on a vehicle with essentially marginal cost added.
“I believe that LIDARs would be very much smaller and much cheaper, like sub $100 and even lower than that, and I have no doubt it will get there,” he says.