One brightest young star in this market is Innoviz, an Israeli start-up founded in 2016 with the aim of developing and commercializing 3D lidar based on MEMS mirror beam steering. It counts Magna International amongst its investors. In the past three years, Innoviz accumulated an impressive $252 million USD and secured a contract with BMW to supply the InnovizOne product in 2021 for a BMW production vehicle. InnovizOne is positioned as an automotive grade MEMS-based lidar module. The question is now whether it will finally gain commercial adoption or not.
A rather new player on the scene is Livox. They are set up by the Chinese drone maker DJI and seem to be following the same strategy: offering a compelling price in order to drive others out and force an industry consolidation. This is a tempting strategy given that there are so many players and technologies today and that most offerings are too expensive and only promise future cost reductions, believe the IDTechEx experts. The lidar is also ToF operating at 905nm. Their Mid-40 lidar with 38.4deg FoV, 0.05-0.1 angular resolution and 260m is using a rotating prism. The interesting point here is that DJI - thanks to its drone business - has excellent know-how and capability in designing the motors, which are often a cost and reliability bottleneck. They have already received a 5k unit order and completed 13k hours of reliability testing running 24/7.
As competition intensifies in the automotive lidar market, it seems likely that some start-ups will not survive. Oryx Vision is the first high-profile casualty: the Israeli 3D lidar business closed in August 2019 despite having secured $67 million USD in funding. Oryx publicly stated delays in the automotive industry as an important factor in the decision to end operations. IDTechEx however believe that their unusual and complex lidar architecture was the main reason. Oryx Vision described their lidar design as 'coherent flash' technology; it operates at a wavelength of 10 micrometres that common and cost-effective lasers cannot produce.
The report also examines the companies that already succeeded in securing a contract with a global car manufacturer. Major automotive OEMs prefer to obtain lidar technology for their autonomous production vehicles via trusted Tier 1 partners. The general trend in the automotive industry is that Tier 2 suppliers develop new lidar technologies, while Tier 1 suppliers optimize these products for high-volume production. For example, Valeo Scala technology was originally developed by Ibeo Automotive Systems GmbH. Valeo developed Ibeo's lidar into a mass-produced sensor for Audi. Similarly, Innoviz has partnered with Magna, the Canadian Tier 1 supplier, and Magna will deliver automotive grade lidar modules to BMW.
As a separate note that Ibeo is now developing a 905nm ToF lidar using SPAD detectors which require multiple measurements to build up the picture but give out a digital signal. Their design includes a VCSEL array and an SPAD array (12800 pixels and emitters). They have three ASICs which allow them to control the system and generate specific patterns on the VCSEL without the need to flash the entire scene. That enables them to boost the range to 250m even though the system architecture and components are similar to a flash lidar. The IC placement and assembly are also relatively straight forward allowing high level of automation. In general, this puts them on a good cost-reduction path which may mean that the lack of a scanning mechanism could be compensated by multiple lidars. Ibeo is working with their investors ZF Friedrichshafen AG.