Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Sub-$200 Lidar could reshuffle auto sensor economics

spectrum.ieee.org

Post reply on HN
Print view View original

361–370 of 582 posts

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#361

Post by vardump »

KaiserPro wrote: post #55
small_model wrote: post #23

Earlier quoted context omitted.

Yes, silly using just cameras, I mean humans have Lidar sensors, that why they can drive, why didn't new just copy that....oh wait. It all seriousness though, Tesla are producing cyber cabs now which are 10th the price of Waymo's and can drive autonomously anywhere in the world. I think we can see where this is going. (Hint: not well for Waymo) Also the article is speculative 'MicroVision says its sensor could one da…

> Tesla are producing cyber cabs now which are 10th the price of Waymo's and can drive autonomously anywhere in the world. Wait what? when did they actually enter mass production? > I mean humans have Lidar sensors Real time slam is actually pretty good, the hard part is reliable object detection using just vision. Tesla's forward facing cameras are effectively monocular, which means that its much much harder to get…

> Tesla's forward facing cameras are effectively monocular

Notably, human perception is effectively monocular in driving situations at distances of 60 feet or farther. It's best in the area where your limbs can reach.

We don't need stereoscopic vision to drive.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#362

Post by aix1 »

pu_pe wrote: post #24
oblio wrote: post #11

Earlier quoted context omitted.

And, I guess, even more advanced surveillance.

LIDAR would be preferrable to cameras when it comes to privacy actually

> LIDAR would be preferrable to cameras when it comes to privacy actually

Right, but how likely is it that there will be LIDAR and no cameras (especially given the low cost of the latter)?

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#363

Post by zamadatix »

b112 wrote: post #165
hinoki wrote: post #87

Earlier quoted context omitted.

I had to look this up, because I had never heard of it. How could a lens be damaged by infrared lasers? It turns out it’s the sensors that are easily damaged by high powered lidar lasers. https://spectrum.ieee.org/amp/keeping-lidars-from-zapping-ca...

If this is true, the eyes are no better. Especially as it can't be seen, who will look awsy? And at night, with open irises? There was someone who had his eyes damaged by sitting next to a heater.

Are the eyes really "no better" in this scenario? From the above article it seems we tuned the behavior to the eye specifically (but not necessarily image sensors):

> Moving to a longer wavelength that does not penetrate the human eye allows new lidars to fire more powerful pulses and stretch their range beyond 200 meters, far enough for stopping faster cars. Now a claim of lidar damage to the charge-coupled-device (CCD) sensor on a photographer's electronic camera has raised concern that new eye-safe long-wavelength lidars might endanger electronic eyes.

> Producers of laser light shows are well aware that laser beams can damage electronic eyes. “Camera sensors are, in general, more susceptible to damage than the human eye,” warns the International Laser Display Association

"doesn't penetrate the human eye" seems a bit hand wavy, but I take it to mean "these length pulses in this wavelength are tuned to have the power not be enough to damage the eye". Camera lenses may not have the same level of IR filtering/gathering area or, if they do, there is nothing implying the image sensor has the exact same tolerances as the inside of the eye. From the same:

> Sensor vulnerability to infrared damage would depend on the design of the infrared filters

A heater usually damages the eyes through drying out/heating up the outside layer with constant high intensity, not by causing damage to the retina (post filtering). https://hps.org/publicinformation/ate/q12691/

> Furthermore, since the eye blocks the IRR, the eye begins to overheat leading to eye damage and possible blindness. Because of this, you should not look at the heater for an extended period of time.

Enough intensity of any wavelength is enough to damage any camera or eye of course, but the scenario here seems to be built around that question for the eye. Similarly, I've heard of Waymo's causing 6 mph accidents but no reports of eye damage from any car LiDAR. Despite that, in the above YouTube clip Marques Brownlee actively shows his camera being clearly damaged as its moved around.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#364

Post by sideband »

xp84 wrote: post #306
Zigurd wrote: post #268

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…

I’m not disagreeing with what you’re saying, but does Alphabet actually intend Waymo to be a trillion dollar retail car business itself, selling cars to everyone? Or would they be happy to sell all those super cool things to OEMs? In a world where “everyone” can make a car affordable that can run Waymo’s software, they may be happy to license all that to “everyone” and simply collect fat royalty checks, à la Microsof…

Alphabet wants drivers on their devices looking at ads instead of driving.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#365

Post by bastawhiz »

torginus wrote: post #346

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…

I'm not sure anyone today really thinks self driving hinges on the hardware. Comma does a surprisingly good job with very minimal hardware (in the form factor of an old Tom Tom!). The advantage is really the device's processing power (cramming enough compute in without making it crazy expensive) and the data that the manufacturer has about the environment and training data to handle edge cases. You can't just buy those things, because the people that have them would be your competitors.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#366

Post by estearum »

kevlened wrote: post #289
estearum wrote: post #159

Earlier quoted context omitted.

I don't think that's the reasoning. The reasoning was simply that LIDAR was (and incorrectly predicted to always be) significantly more expensive than cameras, and hypothetically that should be fine because, well, humans drive with only two eyes. Musk miscalculated on 1) cost reduction in LIDAR and 2) how incredible the human brain is compared to computers. Having similar sensors certainly doesn't guarantee your acci…

> Musk miscalculated on 1) cost reduction in LIDAR Given that Musk has a history of driving lower costs, it's unlikely he overestimated the long-term cost floor. He just thought we were close to self-driving in 2014. Another factor is Andrej Karpathy, who was the primary architect for the vision-only approach. Musk wanted fewer parts, and Karpathy believed he could deliver that. Karpathy is still an advocate of visio…

Right, for the reasons that I just mentioned

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#367

Post by jayd16 »

rbbydotdev wrote: post #351

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?

They never used LiDAR. They removed the radar.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#368

Post by briandw »

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.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#369

Post by vardump »

disillusioned wrote: post #37
small_model wrote: post #23

Earlier quoted context omitted.

Yes, silly using just cameras, I mean humans have Lidar sensors, that why they can drive, why didn't new just copy that....oh wait. It all seriousness though, Tesla are producing cyber cabs now which are 10th the price of Waymo's and can drive autonomously anywhere in the world. I think we can see where this is going. (Hint: not well for Waymo) Also the article is speculative 'MicroVision says its sensor could one da…

This is a weirdly tired counterpoint that Elon and Elonstans like to bandy about as if it's an apples to apples comparison. Humans have a weirdly ultra-high-dynamic-range binocular vision system mounted on an advanced ptz/swivel gimbal that allows for a great degree of freedom of movement, parallax effects, and a complex heuristic system for analyzing vision data. The Tesla FSD system has... well, sure, a few more ca…

> My Tesla has no clue at all that that's happening. And worse, the perpendicular camera responsible for checking cross-traffic is mounted _behind my head_ on the b-pillar

It has a wide angle camera in front that you usually can never see outside service menu. It should cover that case.

Re: Sub-$200 Lidar could reshuffle auto sensor economics

#370

Post by StevenNunez »

Zigurd wrote: post #268

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…

Time to extend comma.ai!
Post reply on HN