Auto Futures first interviewed Omer Keilaf, CEO and co-founder of Innoviz Technologies, two years ago at the Paris Motor Show. As we all know, a lot has changed since then. So we thought it was well worth catching-up with him to get an update on the Israeli start-up that’s now a leading provider of high-performance, solid-state LiDAR sensors and perception software.
It’s been an exhilarating two years for Keilaf and his team. It’s established partnerships with world leading Tier-1 automotive suppliers HARMAN and China’s HiRain Technologies, expanded into four new regions, moved to new offices, and closed a Series C funding round, bringing its total funding to $251 million.
Major developments have included R&D for its LiDAR sensors, InnovizOne and the recently released InnovizTwo, and enhancing its perception software capabilities. Keilaf tells us that its next generation product, InnovizTwo, solves a significant bottleneck in the industry.
“Our technology enables car manufacturers and Tier-1s to offer safe Level 2+, while paving the way to full Level 3 automation. While Level 2+ drivers are still liable for the cars they drive, more and more data is collected while driving, and more and more features will continue to be validated. This will lead to the safety being statistically proven, following which the software can be upgraded to full Level 3 without any hardware change.”
“InnovizTwo is a scanning LiDAR using our proprietary MEMS technology. It meets the market demand of 200m at 10% reflectivity at its highest resolution of 0.1 x 0.1, and it will be smaller, higher performing, and significantly less power consuming than InnovizOne. The 70% lower cost is very important because we’ve pushed the cost of LiDAR down to a level at which more automakers are going to be willing to commit to Level 2+ on a mass market scale,” he explains.
The Growing Demand for Sensors
Three years ago, BMW selected InnovizOne for series production of its autonomous vehicles. Since then, Innoviz has been working closely with BMW, to constantly improve its high-end LiDAR for BMW’s autonomous driving programme.
Alejandro Vukotich, former Senior VP of Automated Driving and Driver Assistance Systems at Audi and BMW comments: “Aside from all the technical challenges, it is key to achieve a price point that enables the industry to introduce a commercially viable technology in the market as a L2 system, and then to incrementally validate and verify the L3 in the field before it gets released as a software update. The InnovizTwo is definitely an important and huge milestone in this direction.”
In March, 2019, Innoviz announced that it had raised $132 million in Series C funding. The funding has helped the Israeli company to keep on expanding its international presence in key automotive markets including Germany, North America, Japan and China.
“Our funding is supporting several key initiatives, including enhancing our LiDAR hardware offerings and our perception software that remain a significant differentiator for us and recruiting the industry’s top talent to enable us to reach our goals.
“It’s also being used to continue our broader business activities, including accelerating our path to mass production and the commercialization of our solutions that address the growing global demand for sensing solutions that enable autonomy,” adds Keilaf.
A big change will be the transition of the driving liability from the driver to the car manufacturer.
Obviously, there are still challenges ahead for the safe deployment of autonomous vehicles (AVs) on city streets.
Keilaf explains: “Automakers still need to log millions of kilometers with functional Level 2+ vehicles (full self-driving with the constant ability and expectation that a human driver takes control) to reach the proper level of confidence in offering even this level of autonomous mobility to drivers.”
Elon Musk has famously stated that LiDAR sensors are too expensive and unnecessary. Tesla relies on cameras and radar for its assisted driving features. But Keilaf says the major players in the car industry are in agreement that no sensor alone can drive an AV.
“Safety is the number one priority in enabling the autonomous revolution, and safety requirements demand a blend of complementary sensors, including LiDAR, cameras, and radar. LiDAR, which is evolving rapidly, already enables better object detection using its 3D imaging capabilities. These capabilities become even more crucial at night, as well as in poor light and weather conditions, seeing what other sensors miss,” he says.
With driverless technology developing at high-speed, we asked Keilaf what urban mobility will look like by the end of the decade.
He tells us: “We anticipate that major automakers will deliver on their Level 3 AV promises over the next ten years. So urban mobility will be highly automated by 2030, and a big change will be the transition of the driving liability from the driver to the car manufacturer. This will most likely have implications on the insurance sector as well.”
“Robotaxis will also be a big piece of urban mobility, starting with applications such as shuttles in geofenced areas, truck platooning, and traffic jam management on highways,” predicts Keilaf.