Featured LIDAR business practices

You missed the part where the drone is restricted to three feet from the ground.

 

I think the whole thing is that some rich guy makes propaganda about how his product is better. Then people who follow the guy or his company become spokespeople for said company and promote that company's ideas and products.

In the end, we don't know yet what will "win" the technological war in the following years. If Teslas didn't have any accidents while using full-self-driving, then there'd be a case. Promises that the future will be this or that are used to win over more investors, but at the same time are just words that mean nothing until the product is actually presented. I'm still waiting on the second gen Tesla Roadster with pneumatic rockets, the $39,900 Cyber Truck, the $25,000 "Model 2" as well as any Tesla I can just jump in the back seat and let it drive me to wherever I need to go.

 

That's a very good analogy. Who would have thought that we'd transition to LCD and OLED? Even by today's standards, CRT's have certain advantages over all flat-panel type of technologies. Image and frame response is perfect with zero delay or latency in a CRT, whereas even the best micro-led still have some latency. CRT's have the absolute best black frame insertion (BFI) over any other technology. And their resolutions and pixel sizes are scalable, especially on tri-tube CRT projectors. But yet, nobody manufactures CRTs anymore, and very few people buy them.

It's not a matter of what is best "in my opinion" because if we used what is best in my opinion we'd all be using CRTs still and if we all used what's best in Elon Musk's opinion then Waymo would ditch their LIDAR systems for "superior" camera tracking technology. Personally, I'll believe it when I see it. Either one could be like plasma TV, that had a rather short commercial life span after being touted as "the best in display technology."

 

Tesla drivers are responsble to supervise the level 2+ ADAS. In the US I believe there are only restricted systems above this level.

Being in a waymo, robotaxi, and 3 other autonomous systems, I really dislike the number of waymos on the road in my city. They are constantly causing traffic jams. They keep passing school buses with their stop signs on, so much so that they will soon be restricted from those routes while school is in session. Now they have been allowed to go on the highway. I wish there were fewer until they get this right.

Years ago I had one of those expensive large CRTs. I had to make adjustments for the magnetic field of the earth, and it was very heavy. Go large and the economics for crts gets worse. This is not the same for cameras and radar that are both small and costs have come down. They scale.

Waymo has a full camera system and radar. The Lidiar is an additonal item. Like the CRT waymo uses a mechanical system that spins. I would say those are more of a proof of concept and latest estimate of the 4 or 5 lidar sensors are about $10,000. That explains why tesla doesn't want to put them on their cars. Now multiple companies are working on mems systems, they use an oscilating mirror. These promise to be much cheaper, lighter, and less power hungry. I would bet that tesla will copy someone if they use it first or maybe are working on this now. Since most lidar companies have gone bankrupt there has been consoidtation but market could be quite large.

There are two very different approaches. Waymo use the best sensors (now made in house) and work hard to get it to work. Tesla build an affordable suite of sensors and have the cars collect the data to get it right. My opinion from riding in both my car and waymo the tesla is closer to working but lidar and radar and more compute power will help. Waymo is working to make the system better and bring costs down. Both have a ways to go to not need a safety driver in many conditions.

 

Size was not part of the analogy. How tech and its price can suddenly change is the analogy.

How can anyone, at this moment, tell everyone else, what is going to be the most effective and economic autonomous driving system 10 or 20 years from now? How do we know that LIDAR will cost $10,000 per car 10 years from now? Sure, cameras are cheap, but what if LIDAR comes down to a couple hundred dollars, fit in a box big enough to hide in front of a rearview mirror, yet can supply data way more valuable than a bunch of cameras?

Again, going back to CRTs. Why do we now use LCDs and OLEDS when CRTs were so much cheaper and better? I don't care about size, size has nothing to do with the point here. Why do we use LCDs and OLEDs today if at one point these types of monitor displays used to cost tens of thousands of dollars and were worse than CRTs in every way, shape and form except weight?

 

. . . and forgo the opportunity to prevent far more deaths and serious injuries on our streets and highways than can be done with merely "good enough"?

The self-driving goal is turning out to be much more difficult than expected or promised a decade ago. We should not be pre-hobbling the solution space by preemptively dismissing whole fields of technology.

 

Actually I was agreeing with you on that point. If we correctly use the CRT to flat panel analogy, crt's are relatively big and heavy for large sizes, but proved you could display great computer images and documents. Then we get to flat panels that allowed for bigger, lighter cheaper. I prius mfd was a small monocrom crt, gen II moved to color flat panel display, now cars have relatively big flat panel displays.

Well with lidar we already have some of these in terms of oscilating mirrors with arrays of sensorrs that are much cheaper. Some claim as low as $200/sensor. Now to get the coverage of the rotating sensors on a waymo you would need more of these smaller cheaper sensors, but I think you don't need that full coverage. Maybe 3 lidar, 3 radar, and 8 cameras would do it. Currently waymo uses 6 radar, 13 cameras and 4 or 5 lidar in the sixth generation. AVEA who mercedes is using in its NVDIA ADAS system say prices are less than $500/sensor and AVEA says you need 2 for ADAS 2+ which is what mercedes is making, and 3-4 for robotaxi operation. That drops prices with today's tech from $10,000 to less than $2000.

I told you why I use them, they are more dependable and less expensive ;-). I wasn't arguing against your point. I just was saying at today's prices there is a reason tesla would not be using the expensive large, heavy waymo lidar. Totally think high end cars will use lidar in the future.