Test & Measurement

Lidar's future in sensing for autonomous vehicles is not assured; enter dot-projector technology by Terranet

Jan. 10, 2020

Lidar as a main sensor for autonomous vehicles (AVs) is still too expensive; cameras could work, but are there any other potential optical sensing technologies for AVs?

John Wallace

Terranet says its VoxelFlow optical sensing technology for AVs captures moving 3D objects with fast edge detection and zero motion blur. Shown is a still from Terranet's video at https://vimeo.com/377730923.

If a member of the general public (or at least one who follows technical developments) were to be asked what kinds of technologies are being developed as sensors for autonomous vehicles (AVs), probably the first answer would be—lidar. Two generalities concerning the lidar arena are true: a plethora of lidar startups has cropped up, and the hype around lidar is quite pronounced. Yet, the predominance of lidar in AVs of the future is not a sure thing.

Consider what Elon Musk, founder of Tesla Motors and developer of AV technology, who is pushing cameras plus vision software as the sensor solution for AVs, has said: "Lidar is a fool's errand. Anyone relying on lidar is doomed. Doomed!"

Of equal or greater relevance are comments made by presenters at the Optical Society's (OSA's) OIDA 2019 Forum on Optics in Autonomy and Sensing (27 June 2019; San Jose, CA):

Yvonne Lutsch, investment principal at Robert Bosch Venture Capital (a branch of Bosch, a $90 billion company in the mobility sector), advised not to focus on technology as much as to focus on the "magic triangle"—performance, maturity, and cost. She noted that there are more than 70 lidar companies in AV, with not a single hugely successful startup "exit" to be seen. Lidar cost is high, she emphasized.

Alixi Andreev, a physicist who is cofounder and managing director of Autotech Ventures, a venture-capital firm focusing on ground transportation, noted that Homo Sapiens is proof that lidar and radar are not required for AVs.

Anand Joshi, principal analyst/strategy consultant in artificial intelligence at Tractica, said that he didn't think that $500 lidar would work; this would seem to indicate that lidar for other than higher-end cars might not be practical.

Carl Conti, technical director at Spatial Integrated Systems (SIS), which does mostly defense work, said he was "agnostic to sensor type." He projected that autonomy for any vehicle would consist of radar, electro-optic stereo, IR stereo, and only occasionally lidar.

It should be noted that lidar was very definitely on the table in discussions at the OIDA forum; however, among these experts (in contrast with some of what's seen in the general press), lidar is only one of many potential AV technologies. In reality, sensor fusion will unite more than one AV technology; the question is, will lidar be among them?

Enter Terranet

Based on the comments above, it would seem that any new and different AV sensor technologies would be both welcomed and very carefully examined.

At CES 2020 (Las Vegas, NV; Jan. 7 to 10, 2020), a Swedish company called Terranet is demonstrating its own optical, but non-lidar, entry into the world of AV sensor technology, called VoxelFlow. The company describes it as "a 3D motion perception technology capable of detecting and classifying objects far more accurately and exponentially faster than Lidar." Terranet licensed the technology from consultant Gerard "Dirk" Smits.

Terranet further describes the system: "The underlying problem with Lidar or any existing camera/computer vision technology is that it cannot detect vulnerable road users with low latency. VoxelFlow on the other hand classifies dynamic moving objects at extremely low latency using very low computational power, producing 10 million 3D points per second, resulting in rapid edge detection without motion blur. The iPhone's "selfie" camera pales in comparison, producing a mere 33,000 light points per frame. Put into context, standard vehicles today and even those with autonomous driving roadmaps in place don’t use cameras much stronger than that of the iPhone's selfie camera."

The newer iPhones' Face ID feature is based on a vertical-cavity surface-emitting laser (VCSEL) system that projects 33,000 IR dots to capture the scene. From Terranet's description, the company's system is similar at least in the fact that it is based on dot projection—10 million dots for VoxelFlow.

According to Terranet, VoxelFLow is fast: "By providing life-saving decision making in milliseconds versus seconds, Terranet envisions transforming the pre-crash inattention that today accounts for roughly 1.35M road accident deaths into a forgotten collision-free swerve on the road."

Info on Terranet can be found at www.terranet.se. To schedule a VoxelFlow demo at CES, email [email protected]. A video on VoxelFlow uploaded by Terranet to Vimeo can be seen at https://vimeo.com/377730923.

Source: https://terranet.se/press/terranet-to-present-first-serious-lidar-alternative-at-ces/

Got optics- and photonics-related news to share with us? Contact John Wallace, Senior Editor, Laser Focus World

About the Author

John Wallace | Senior Technical Editor (1998-2022)

John Wallace was with Laser Focus World for nearly 25 years, retiring in late June 2022. He obtained a bachelor's degree in mechanical engineering and physics at Rutgers University and a master's in optical engineering at the University of Rochester. Before becoming an editor, John worked as an engineer at RCA, Exxon, Eastman Kodak, and GCA Corporation.