NIO’s investment in Innovusion can be viewed as a strategic move for its foray into autonomous driving, seizing opportunities in the AV supply chain. Yet, for Innovusion, it remains to be seen whether its 300-line LiDAR tech is just a hype.
Innovusion, a 3-year-old startup headquartered in Silicon Valley and specializing in LiDAR (Light Detection and Ranging) sensor technology to enable safe Level 4-5 AV (autonomous vehicles) with volume production, secured a Series A funding of USD 30 million in Oct. 2018.
This round was co-led by China’s Tesla-rival NIO (蔚来) Capital and Eight Roads Ventures, featuring US-based F-Prime Capital, Gaorong Capital and other strategic investors that had provided its seed funding of USD 3 million. It was co-founded by BAO Junwei (鲍君威) who previously led the sensor and hardware team at Baidu’s Autonomous Business Unit, and LI Yimin (李义民) who's a former member in Velodyne’s (威力登) core LiDAR R&D team.
NIO’s investment in Innovusion can be viewed as a strategic move for Nio’s foray into AV, seizing opportunities in the AV industrial chain. Right before this investment, Nio-backed AV maker Momenta received a funding of USD 200 million at a valuation of USD 2 billion.
According to ZHEN Yun (郑赟) with Roland Berger, Level 3 autonomous driving can become a reality in urban cities while highways can find Level 4 self-driving cars from 2026 to 2030. In 2040 and the years to come, all cars on the roads are to be full Level 5 autonomous.
LiDAR, as the “eyes” for self-driving cars to perceive the surroundings and make right and timely decisions, is well-accepted to be a crucial component to the safe and scale roll-out of AV. 93% of autonomous vehicle experts within the field interviewed by UBS (United Bank of Switzerland 瑞银集团) believe that LiDAR is a prerequisite for autonomous vehicles.
Tesla chooses a different path and uses a combination of cameras, ultrasonic sensors, and radars rather than LiDAR units in its self-driving action. Despite Elon Musk insists that LiDAR sensors are too costly and unnecessary for full driverless cars, many car OEMs and tech giants such as Alphabet’s Waymo and Baidu, believe in it and endeavor to embed LiDAR into their autonomous cars, resulting in the soaring need for it.
High price (ranging up to USD 100 thousand) and huge demand, the LiDAR industry has seen start-ups springing up and receiving funding to advance technology development for lower prices and accelerate mass production since 2015, with over USD 1 billion capital pouring into this field. According to Yole Développement, the global market of LiDARs is estimated to reach USD 5 billion in 2023 and quintuple to USD 28 billion in 2032.
The arena is becoming increasingly crowded with over 50 companies, competitive and fragmented, growing substantially at a CAGR (Compound Annual Growth Rate) of over 37%. New players are trying to catch up with the long-established global LiDAR leader, Velodyne (威力登), which locates in Silicon Valley and holds the lion’s share (80%) of the world’s LiDAR sensor device supply to Tier-1 companies and OEMs, at an expected sales volume of over USD 400 million in 2018.
Compared with some other LiDAR tech companies (as shown in the table below), Innovusion’s latest investment is relatively small which raise doubts about its 2017 product Innovusion Cheetah (猎豹) that it claimed to be "the world’s first image-grade LiDAR system" capable of supporting “Level 4+” autonomous driving with near picture quality with a 300-line resolution point cloud to detect objects more than 200 meters away.
However, the company did not specify on how exactly this product was going to support L4 AV with illustrative scenarios nor the types of technology it was based on. It began shipping samples of Cheetah LiDAR in Q2 2018 and ramped up production in Q4. Until now, the common commercially available and practically used LiDAR has been Velodyne’s HDL-64E costing USD 75 thousand. The best Velodyne can offer so far is ALPHA PUCK (VLS-128) which is a 300-meter sensor set to replace HDL-64 when at volume production.
Whether Innovusion’s self-claiming tech can up the ante on existing LiDAR needs to be consolidated through strategic partnerships with Tier 1 companies or OEMs, but little is disclosed regarding its partnerships, even on its official website; while other players have teamed with well-known OEMs such as Daimler AG, Toyota, BMW, Hyundai, etc.
Another challenge for Innovusion is that funding is becoming harder as one Gartner analyst, Mike Ramsey, noted that many LiDAR start-ups were running into technological bottlenecks in terms of the conventional LiDAR tech as well as difficulties in commercialization because some of them failed to guarantee their products’ performance as they were promoted and AVs were not ready for mass production yet. It was reported that the practical test outcomes of Quanergy’s S3, M8 devices were hyped and its previous partner Daimler didn’t list Quanergy in the autonomous taxi project. Many LiDAR makers began turning to consulting service, providing tech advice and guidance.
Additionally, the race to more affordable autonomous cars keeps witnessing new cost-effective, increased-performing solutions for scale production. Earlier this month at CES (Consumer Electronic Show) 2019, Insight LiDAR announced the development of digital coherent LiDAR, a chip-scale, long-range sensor targeted at the emerging autonomous vehicle market. It is based on FMCW (Frequency Modulated Continuous Wave) technology with advantages over the current generation of ToF (Time-of-Flight) LiDAR sensors. Israeli company Innoviz Technologies has been awarded with the 2019 CES Best of Innovation honor for its InnovizOne automotive-grade solid-state LiDAR device with the highest angular resolution of 0.1 x 0.1 degrees.
In Nov. 2018, Ouster revealed its 3D sensing multi-beam flash LiDAR solid-state solution which uses only two chips and 85% cheaper. It was patented and started shipping to some OEM partners of the company. Earlier than that, in July 2018, another LiDAR maker, Australia-based Baraja, launched an innovated system that uses prism-like optics and shifting wavelengths of light called Spectrum-Scan for greater accuracy and more software-/processing-algorithm- compatibility.
LiDAR currently is a seller’s market while innovated at a fast pace, very similar to the path of the widely-used ABS (Anti-skid Brake System): costing USD 8 thousand in 1979 but only USD 20 now. Nevertheless, Innovusion’s growing space is squeezed when so many players are out there, investing in the systems and processes required to manufacture LiDAR at scale. To survive and gain advantage in the market, the key is to accumulate production resources (capital and equipment/components suppliers), integrated other autonomous driving technologies for upgraded configuration and cooperate with AV tech Tier 1 companies, OEMs and other companies that have scenarios of LiDAR application.