Lidar (Light Detection and Ranging) has garnered lot of attention as a remote sensing technology in the fledging field of self-driving vehicles. At the first Autonomous Vehicles Sensors Conference in San Jose Tuesday, several startup companies discussed their approaches to developing lidar technology, in a packed afternoon conference session.
Autonomous vehicle developers are looking closely at lidar because current radar-based sensor technologies are inadequate and have bad angular resolution, says session chairperson Jake Li, Business Development Manager of Auto Lidar at Hamamatsu Corp. Hamamatsu builds emitters, photodetectors, and other essential components for lidar systems. Li expects early adopters of lidar to be fleet vehicles and taxis, predicting consumer vehicle applications perhaps by 2025.
With some help from investor funding, these companies have intense R&D underway to design lidar into future autonomous vehicles. So far, some companies appear to be aiming at more advanced lidar technology, while others are countering with more conventional technology they hope to scale up later on.
Stephen Lambright, Vice President of Marketing for San Francisco-based AEye, described a lidar approach that integrates the company’s iDAR (Intelligent Detection and Ranging) lidar sensor. Earlier this year at the Consumer Electronics Show, AEye unveiled its AE-100 robotic perception system, integrating three core components: reportedly the world’s first agile MOEMS lidar, pre-fused and bore-sighted with a low-light HD video camera; distributed artificial intelligence; and software-definable and extensible hardware. According to Lambright, the use of embedded artificial intelligence pushes some of the processing to the edge of the network rather than concentrate all processing in a central unit.
Taking another approach, Blackmore Sensors and Analytics of Bozeman, Montana, has developed a frequency modulated waveform technology that reportedly provides for more accurate measurements, according to Jim Curry, Vice President of Product. Among other benefits, the technology allows simultaneous measurement of range and velocity, compared to the other lidar techniques that employ pulsed radar techniques, according to Curry.
“There is a high level of interference rejection built in. Our frequency-modulated lidar operates interference-free. There are no issues with sunlight, or adverse effects from other cars’ sensors.”
Getting lidar into production products remains a challenge, but one company claims their technology can do just that. Raffi Mardirosian, Vice President of Corporate Development at Ouster, also located in San Francisco, discussed the company’s OS-1 lidar sensor (see video below). Available now, the sensor uses the same semiconductor technology used in the Apple iPhone X FaceID technology, according to Mardirosian.
The company says the OS-1 is smaller, lighter, uses less power, and is less expensive than existing lidar sensors. It is developing a next-generation lidar sensor with further improvements.
Phantom Intelligence of Quebec City, Canada, is developing what it terms as affordable lidar technology based on a single-chip FPGA technology. “The technology is scalable, is capable of medium resolution, and integrates housekeeping functions into the single-chip FPGA architecture,” says Eric Turenne, Chief Technology Officer. Turenne says the technology would in the future migrate to higher resolution and possibly integrate application intelligence.