Source: Daily Automotive News MICROVISION GETS LIDAR TECH FOR PENNIES ON THE DOLLAR Five years ago the stock prices of EV makers were going through the roof. And the same was largely true of lidar companies. One of those companies, Luminar, saw its stock shoot over $500 a share, giving it an $11 billion market cap, and turning its then 25-year-old founder into a multi-billionaire. But then Luminar went bankrupt and another lidar company, Microvision, just bought all its assets and got most of its customers for only $33 million. In fact, Microvision bought another lidar maker, Scantinel Photonics, after it went bankrupt, too. And Microvision’s CEO, Glen DeVos, says they’ll buy other lidar companies too, if the right situation arises. Maybe you saw the interview we did with DeVos at CES last month. Microvision has developed a $200 lidar and is working on getting the price down to $100. And by buying up bankrupt lidar companies, it’s getting a ton of technology for pennies on the dollar. I have been and remain skeptical of LIDAR because it typically requires mechanical parts moving to scan a volume. Moving parts are the bane of reliability as they inherently wear out and lose calibration. In contrast, I prefer dual camera systems like: no moving parts - once built - no mechanical parts to wear out, deform, or need lubrication. Calibrate once when replaced and run until it breaks. multi-spectral - from infrared through deep blue, it tells the difference between a warm critter and puppet even in the darkest night. This winter, I regularly see in the rendered image, black clad people and animals in the road or on the curb in my Tesla. fast -it scans the volume at the frame rate that ca track flying objects like the rock headed to a windshield. Tesla has been and continues to sell multi-camera, AutoPilot and Full Self Driving systems for nearly a decade, millions of copies. In contrast, LIDAR has been talking about self driving in 3 orders of magnitude fewer cars for almost as long. One delivers and the other is whistling in the dark. So Tuesday February 3, I had an optometrist appointment with full examination of my dilated eyes. A cold, windy rain started when I arrived and by the time I eft, rush hour had begun in earnest. To go home, I sheltered at the patient portal away from the rain and wind. The car drove ~100 to me avoiding other cars and pedestrians. I only had to dash 20 yards to the door. It was rush hour so I directed Full Self Driving (FSD)to take the nearly empty secondary streets to a favorite restaurant of supper.Hands free, I felt no need to criticize the other, manual drivers, no matter how foolish. FSD did not complain either. So I was an interested passenger in a dry and warm car that parked itself when I got there. Driving home, I directed to car to use secondary streets, all but free of traffic and the shortest, direct routes. Hands free, I just had to pay attention to what the car was doing. I parked at a grocery store and once again, the car came to the pickup space on its own. Then home where it charged in a couple of hours while I ate, TV, and took a nap. The ABC of EVs is "Always Be Charging" while doing something you, the human, wants to do. Bob Wilson
Flash/directional/solid-state LIDAR doesn't have moving parts, and has been available for years now. Apple has been using them on their Pro pads and phones since 2020. https://appleinsider.com/articles/20/10/13/apple-launches-iphone-12-pro-line-with-new-design-better-cameras/
obviously from those that have used it given correct hardware and software vision systems are good enough. But this could be better. Waymo's lidar implementation is ugly, noisy, and expensive. A number of companies are working with MEMs type Lidar, these use one or two oscilating mirrors, which means a semi solid state lidar which should be more reliable, smaller, quiet, but less accurate. You need more of these or something else to get a full 360 degree picture. From a development point of view waymo's was a good engineering approach. Use the best lidar and find out what computer hardware and softeware are needed. Because units are small and engineering expensive this made sense. Now proof of concept is out there a combination of mems or true solid state lidar with radar and vision seems the best approach. Hardware and software has improved. Even tesla uses lidar in testing of its systems. I expect in 5 years or so there will be inexpensive enough sensors to put redundancy in. Radar is not as affected by the weather. Lidar is the most accurate with details and depth. Vision helps decide what these things are actually measuring.[/QUOTE]
Yes, have you compared the cross section area of a phase array versus a camera lens? Published LIDAR vehicles have aerodynamic parachutes deployed outside the body - this destroys highway speed efficiency. OK for around town at or under 45 mph but death on the highway. Incorporate at the upper lip of the hood, less bad but that far forward limits what can be detected across the front. bumper impact zone. Already there has been one report of a LIDAR operated taxi rolling over a human on the road. Compare the electronics needed to drive an optical phase array versus existing optical camera lens and image array. LIDAR has a place but the price-performance compared to a digital camera is off by about two orders of magnitude. Bob Wilson
