Blackmore Sensors and Analytics reported its automotive Doppler lidar system for self-driving cars surpassed expectations in real-world tests in San Francisco’s congested streets and on Bay Area highways.
Backed by more than a decade of testing with the U.S. military and supported by investments from BMW i Ventures and Toyota AI Ventures, the Blackmore system has the potential to make autonomous cars much safer than radar or conventional pulsed lidar sensor systems.
“Blackmore’s team is on a mission to make automated driving safer by leveraging FMCW technology to bring Doppler capabilities to the lidar market, a potential game-changer for the autonomous mobility market,” said Jim Adler, managing director of Toyota AI Ventures.
There is no such thing as “too safe” with autonomous vehicles. Any time a self-driving car is involved in an accident, the bar rises for greater safety assurances. Human factors matter when drivers are supposed to be alert to take the wheel when needed, and self-driving algorithms must be fast and accurate, but without timely and precise sensor data, attentive humans and ingenious algorithms won’t keep our roadways safe.
Blackmore uses frequency-modulated continuous wave (FMCW) lidar. Lidar measures distance and speed by bouncing light waves off objects. Lidar is more accurate and faster than radar, which bounces radio waves, but radar has no moving parts and is less expensive than lidar. New developments in solid-state lidar, however, have reduced cost and size. The first autonomous vehicle sensor systems used radar in combination with cameras. Today most car autonomy experts think systems with lidar are required for full autonomy, typically in sensor systems that include lidar, radar, and cameras.
But all lidar is not alike. There are key differences between pulsed and continuous waves which are significant for autonomous driving. Specifically, measuring speed by calculating multiple pulse bounces is slower and more error-prone than detecting the Doppler effect frequency change with continuous waves.
“Traditional pulsed lidar systems do not measure motion directly, meaning velocity information is only available after running computationally complex and error-prone analytics algorithms on multiple frames of data,” according to Blackmore’s news release.
Pulsed lidar is a more mature technology, but FMCW lidar can detect Doppler frequency shifts for near instant velocity or speed detection. Doppler shifts aren’t a new discovery. Austrian physicist Christian Doppler first described the effect named after him in 1842. The Doppler shift, or Doppler effect, is the frequency change measured by wavelength when objects move closer or farther away. Wave frequency increases as objects get closer, and the frequency decreases as objects get further away.
“Engineers with leading autonomous driving teams are ecstatic to see Blackmore’s Doppler lidar in action,” said Randy Reibel, Blackmore’s CEO. “They understand the value of Blackmore’s unprecedented approach and are excited to tap into a system that is economically efficient, reliable, and scalable, without compromise.”
