Tesla is adding radar and I predict before long it will add LiDAR because that's the only way to get to Level 3, which is a requirement for moving forward in California

https://www.fccidlookup.com/report/tesla-new-millimeter-wave...

I worked in mobile robotics for a defense contractor in the early-mid 2000s and we had a homegrown Lidar (though we always called it ladar) that was large, heavy, and cost $250k to make. I remember a year out of college driving 2 hours to a military base with it in the backseat of my car and being paranoid at every bump I hit. These things seem like a dream.

Since lidar has distance information and cameras do not, it was always a ridiculous idea by a certain company to use cameras only. Lidar using cars are going to replace at least the ones that don't make use of this obvious answer to obstacle detection challenges.

Before y'all say that now everyone will be able to get Waymo's sensor suite for hundreds of dollars instead of tens of thousands, that's the easy part.

Waymo benefits from Google's unparalleled geospatial data. Waymo also has a support architecture that doesn't depend on real time remote operation, which can't be implemented reliably in almost all cases. You can't be following your supposedly unsupervised cars with a supervisor in a chase car. You can't even be driving remotely. Your driver software has to be able to drive independently in all cases, even those where it needs to ask a human how to proceed.

The difference between level two and level three driver assist and level four autonomy is like the difference between suborbital flight and putting a payload in orbit. What looks like a next logical step actually takes 10X or more effort, scale, and testing.

This $200 MicroVision lidar is a short range lidar that produces a really fuzzy point cloud. At best, it can be used for parking. It's unlikely to help self driving cars much at all, much less "reshuffle auto sensor economics".

I've always wondered if Tesla's issues with FSD were a sensor problem or an intelligence problem. I think Tesla's claim is that when they look at accident footage, it is clear to a human how the car could have avoided the accident, and thus, if FSD was more intelligent, the accident could have been avoided. Is this reasoning wrong?

I personally find it convincing that the problem with self-driving is mostly that the models aren't intelligent enough, and that adding LiDAR wouldn't be enough to achieve the reliability required. But I don't know, I don't really work in that field so maybe engineers who have more experience with self driving might say otherwise.

The mind salivates at the idea of sub-$100 and soon after sub-$10 Lidar. We could build spatial awareness into damn near everything. It'll be a cambrian explosion of autonomous robots.

Low cost, sub $200 automotive grade LIDAR sensors are already available.

Cepton Technologies offers Nova [0], Nova-Ultra [1] sensors both at a sub-$100 price point [2]. These feature a 120°(H) x 90°(V) FOV at 50m, with 2.7M points per second sampling.

Velodyne introduced Velabit in 2021, for $100. Boasting 100m range and a 60-degree horizontal FoV x 10-degree vertical FoV.

The article claims that:

> What distinguishes current claims is the explicit focus on sub-$200 pricing tied to production volume rather than future prototypes or limited pilot runs.

which is simply not true. Cepton (currently offering) and Velodyne (acquired by Ouster in 2023) have done this for years.

  [0]: https://www.cepton.com/products/nova
  [1]: https://www.cepton.com/products/nova-ultra
  [2]: https://www.cepton.com/announcements/ceptons-nova-lidar-named-as-ces-2022-innovation-awards-honoree
  [3]: https://lidarmag.com/2020/01/07/velodyne-lidar-introduces-velabit/

Interestingly, there have been people in the LIDAR industry predicting costs like this for many years. I heard numbers like $250 per vehicle back in 2012 [1]

Of course, ambitious pricing like this is all about economies of scale - sensors that are used in production vehicles are ordered by the million, and that lowers the costs massively. When the huge orders didn't materialise, the economies of scale and low prices didn't materialise either.

[1] https://web.archive.org/web/20161013165833/http://content.us...

'MicroVision says its sensor could one day break the $100 barrier'. When an article says one day, read not in the next decade.

I wonder if Comma.ai will ever be open to incorporating this into openpilot.

I always thought the argument that humans are adequate drivers and hence only cameras was not great. Why not actually be better than humans at sensing and driving?

Interesting to see the cost curve drop ... this always changes the market.

I have been watching the sensor space for a while. Cheap LIDAR units could open up weird DIY uses and not just cars. ALSO regulatory and mapping integration will matter. I tried to work with public datasets and it's messy. The hardware is only one part! BUT it's exciting to see multiple vendors in the space. Competition might push vendors to refine the software stack as well as the hardware. HOWEVER I'm keeping an eye on how these systems handle edge cases in bad weather. I don't think we have seen enough data yet...

Microvision has been saying that from half a decade, products? Nowhere to be found.

So tiring to keep hearing this argument "humans only use vision to drive, so why would self driving cars need more?"

This argument is inherently anti-progress. It's like saying human had been using sextants to navigate for hundreds of years, why GPS?

A more sensible question is, why not?

> laser pulses

> phased-array

I'm not well versed into RF physics. I had the feeling that light-wave coherency in lasers had to be created at a single source (or amplified as it passes by). That's the first time I hear about phased-array lasers.

Can someone knowledgeable chime in on this?

Are we sure these things aren’t damaging our eyes? It’s lasers shooting all over the place right?

There are laser measurers sold for a few buck on Temu. Robot vacuums sold for few hundred dollars have Lidars that map out the room in a seconds.

Is there any actual technical reason why automobile Lidar be expensive? Just combine visual processing with single point sampler that will feed points of interest and accurate model of the surroundings will be built.

@dang .... do these comments seem organic to you? old accounts with almost zero karma going out of their way to use the same verbiage to compliment waymo 18 minutes after an article gets posted? .... dead internet at work.

From the article:

> pricing below US $200. That’s less than half of typical prices now, and it’s not even the full extent of the company’s ambition.

This means there are sensors available for like $500 or more. At 4 per car, this is still just $2000, which is a very reasonable cost add even for a midrange car.

And with price comparisons like this, I'm sure Chinese competitors aren't factored in, I'm sure the Chinese have stuff for cheaper.

So Affordable Lidar is not a limitation. Despite that, self-driving doesn't really exist outside of Waymo, which people take to assume that Lidar is their killer advantage, but with other cars having Lidar, I think that might not turn out to be the deciding facotr.

When every car has LIDAR will they all begin to blind each other?

(Insert old man rant “Why are everyone’s headlights so gosh darn bright these days?!”)

The article is a bit muddy on what is hope and what is product. Can we _really_ buy a solid state lidar today? At what cost? When can I have it delivered?

The article starts out without saying it but my takeaway at the end is "Not $200" and "Not in the near future"?

I never understood why Tesla HAD to get rid of the Lidars. Expensive today sure, but can you imagine all that training data they missed out on? Technology has a way of becoming cheaper and cheaper. It seemed short sighted, even if at a loss, again, the training data.

If the pros of having a camera are monumental, then couldn't the video and lidar be combined to be even greater?

I really wish that companies would just sell their products instead of doing the business relationship 2-step. It is an unnecessary waste of time to sell product.

It looks like these sensors have just enough range to be effective for lidar terrain scanning. I would have bought a Movia S right now just to try it out.

Cameras alone can handle the vast majority of nominal driving scenarios, but the long tail of safety critical edge cases is where progress slows dramatically. Many of these cases are driven by degraded or ambiguous perception, which is where multi‑modal sensing, such as combining cameras with lidar, can reduce uncertainty. In adverse weather like fog or heavy rain, that reduction in uncertainty can translate directly into safer behavior, such as earlier and more confident emergency braking, even if no single sensor performs perfectly on its own

Laser safety people: how concerned should we be about city streets full of aggressively cost-engineered Lidar emitters?

Just to be clear, this article is talking about the possibility that this might happen one day. LiDAR remains prohibitively expensive AFAICT.

I still believe in Cameras. I have a comma.ai 3x and it works really well. Just get a thermal camera to deal with fog etc. Waymo has some of the same limitations with cameras that Comma and Tesla does.

Anyone know what the ballpark total marginal cost to a consumer for increasing the BOM of a car for $1 is?

Radar is extremely expensive, and lifar is just below that.

Glad to see someone lowering the cost of this technology, and hope to see lots of engineers using this tech as a result.

We might even see a boom in LIDAR tech as a result

I wonder if this could be adapted to the vtuber market. Saw a vtuber body tracker being marketed at $11k recently.

Below is one of the comments poster to original article, reading it makes me think that most of the whole article has been regurgitated by some AI:

>"This misleading article contains numerous factual errors regarding automotive lidar. Here are the most glaring:

There are multiple manufacturers, including Hesai, that use mechanical means for at least one scan axis and are already sold for a fraction of the "$10k - $20k" price noted by the author. Luminar itself built this class of scanners before going bankrupt.

Per Microvision's own website, the Movia-S does not use a phased array and also does not have a range anywhere near 200m.

Velodyne and Luminar do not even exist as companies anymore. Both have gone bankrupt and been acquired by competitors."

Would it be no easier to integrate it into home vaccuums ?

Is this Human safe at these volumes? There was a time you could get your feet sized by putting them into an X-ray box at the shoe store. Removed from stores once the harm was known.

What is this author even doing with these numbers?

can I buy it on digikey yet?

BTW what happens when there are hundreds of Lidar signals at one intersection?

There's no way a sensor can tell if a signal was from its origin?

Guessing any signal should be treated as untrusted until verified but I suspect coders won't be doing that unless it's easy

I saw a Waymo in Seattle, today. If Waymo can get Seattle right, that gives me a lot of confidence that their stack is very capable of difficult road conditions.

Note: I have not had the pleasure of riding in one yet, but from what my friend in SJ says, it’s very convenient and confidence-inspiring.

What? You get Chinese lidar sensors for 12 EUR for a long time already.

How could I buy one?

It might, but comma.ai proves that lidar is red herring, which is further supported by the fact that Waymo are able to drive vision-only if necessary.

will Musk backtrack on the whole CV enough, that's how humans do it if price becomes this low?

[dead]

[flagged]

[dead]

Oh hell yeah, we can finally stop the braindead attempts to make a safe self-driving car with just cameras.

The brains (ai models) are more important than the sensors. Cameras are good enough. Lidar doesn’t keep Waymos from driving into an 18” deep puddle, or driving the wrong way down the street. Lidar doesn’t help predict when a pedestrian is going to try to cross the street. Lidar doesn’t give the car the common sense to slow down because a child just ran behind a parked car and will soon be coming out the other side.