Slashdot Mirror
GM Exec Says Elon Musk's Self-Driving Car Claims Are 'Full of Crap' (smh.com.au)
An anonymous reader quotes the Sydney Morning Herald:
Billionaire entrepreneur Elon Musk's claims about the self-driving capabilities of his upcoming Tesla vehicles are "full of crap", General Motors' self-driving Tsar says... "To think you can see everything you need for a level five autonomous car [full self-driving] with cameras and radar, I don't know how you do that"... GM's own solution involves several radar and Lidar sensors, as well as cameras and multiple redundancy systems. Each system costs hundreds of thousands of dollars, and GM are some way away from getting the cost low enough to be commercially viable. "The level of technology and knowing what it takes to do the mission, to say you can be a full level five with just cameras and radars is not physically possible," Mr Miller said.
GM can't do it.
I definitely trust GM to make an unbiased analysis of competitor technological capacity.
I don't possess radar, LiDAR, or a gazillion redundancy systems. Stereo cameras (eyes) on a pivoting head and two directional microphones (ears). My software is way better than GM's, though, and I'm expecting Tesla's is too.
Then why can't computers. We have an advantage that we can fill missing pieces in stereoscopic vision to complete our perception and it is incredibly difficult for algorithms to do that. Radar makes up for that, but it is simply possible that Tesla is closer to that then GM is.
But to clarify the difference:
LIDAR: Formerly about $75k, now about $7,5k per unit, and requires a bubble dome on top of the vehicle. GM and Waymo use it, Tesla doesn't. In addition to looking weird and adding drag, the price is killer if you want to include something on every vehicle. Beyond this, LIDAR doesn't work in fog, heavy rain, snow, and other conditions that humans can drive in - meaning that you'd have to either prevent trips during these conditions, require humans to drive during them, find workarounds (not easy), or rely instead on other sensors. And you still need to understand the world around you visually - LIDAR will tell you that "something" is there, but it can't read signs, see road markings, see brake lights, tell if that thing in the road is a person or a paper bag, etc.
Tesla, for these reasons, ruled out LIDAR. They just simply use the "other sensors" - 1x radar, many cameras, many ultrasonic sensors - all the time. This way, all of their sensors can be put in all of their vehicles, and do double duty for both self driving and standard safety features (autobraking, etc), depending on what options the buyer has paid for. This however comes at a penality: when LIDAR works, it works really well. Photogrammetry with cameras is prone to stitching errors, and radar, while being able to see some things that humans can't, sees the world in very strange ways (for example, a piece of plywood is transparent, but an aluminum can glows like it's on fire). It's a much more challenging task if you leave LIDAR out of the loop. But, it gives you a more saleable product.
In the end, I expect a convergence to take hold. An interesting new technology for example is time-of-flight cameras - they function as normal cameras, but also can read the length of time it takes for a laser pulse to return on every pixel they record. So no dome, just your normal camera coverage and a few cheap, fixed lasers - in mass production, it might not cost much more than cameras alone. In such a case, I'd expect the LIDAR groups to simply replace their conventional LIDAR datastream with the time-of-flight datastreams, while I'd expect the non-LIDAR groups to replace their photogrammetry-and-radar built 3d models with time-of-flight 3d models. But both sides will still need image processing, so it's important to work on maturing that technology today.
That said, let me reiterate that I'm a pessimist regardless of what tech you use. There's just so much nuance in driving in hazardous conditions - understanding when, where and how much you have to slow down, what's safe to drive on and what's not, what things to the side of the road are hazardous and what aren't, when you should break rules (such as driving in the middle of the road when conditions are dangerous but oncoming traffic is rare), what are the consequences of a mistake in one location vs. another, etc, etc. On my gravel road, there's a canyon to one side with no guardrail, and varying amounts of ice and potholes in different places. You better well know how your traction is going to fare as you move across the potholes (vibrating the car and making it lose traction) or icing if you don't want to end up in an unrecoverable slide into a ravine.
Just to pick a random example among countless things that you have to take into account: how long do you think before any self-driving systems will have "sheep recognizing algorithms"? Because where I live, there's sheep. Group of sheep on one side of the road: probably safe. Group of sheep on both sides of the road: not as safe, but probably safe. Lamb on one side, ewe on the other? Very dangerous - the lamb will invariably run to its mother as you approach. Where's the ewe-lamb-running algorithm?
"If there was an antonym to 'Elon Musk', it would be 'Richard Branson'."
Also, as to answer the question of how humans do it with "two cameras": logic. We don't have "stitching errors" in how we build up a model of the world around us from visual data because our brain constantly processes everything around us through the prism of "does that make sense?" But whether something "makes sense" or not is an AI-hard problem.
Building up 3d models with photogrammetry is an inherently error-prone process because a computer doesn't know if something makes sense. The approach is "Oh hey, these patterns from the left and right camera matched up - there's an object there at X distance based on how far the patterns had to be shifted to align". But what if the patterns happen to be different things that just happened to match up in patterning? Or what if, due to lighting / texturing / obstruction / material issues, the same thing didn't look exactly the same from different angles? Uncovering these problems is, as mentioned, AI-hard. I doubt anyone is even trying at this point; there's enough to work on just to get things to follow road lines correctly and not go chasing old tire tracks or poorly erased construction lines.
Ranging systems like LIDAR help let you just simply ignore the issue by telling you flat out, "I sent out a beam in this precise direction and got a reflection in precisely this length of time." But LIDAR has a number of problems, as described above.
"If there was an antonym to 'Elon Musk', it would be 'Richard Branson'."
Also, we crash more often than is permissible to a self-driving vehicle.
Next you're gonna be telling me how we're not all going to have villas on Mars by the end of the next decade, at the very latest, all brought to you by SpaceX brand rocket ships.
Elon gets my respect for making two successful and innovating businesses that have lasted and have solid fundamentals into the future. He gets no credit for his bullshit factory. Good bullshit has to be believable. Self-driving cars and 800 mph trains in a tube by next year doesn't pass the giggle test.
What a stupid thing to say. What we need is a car company that can actually deliver what it promises and while doing so remain economically profitable and not need billion dollar handouts every 5-10 years like GM. When the price comes down, as it has been doing so with each new model of Tesla released, the general public will be able to afford them
Elon's had a long history of proving naysayers like this wrong. My money's (literally) on him to pull this off. The folks truly terrified of self-driving cars are the National Auto Dealers Association (NADA) who stand to lose the most when you can order a car online and have it deliver itself.
Then so can a computer. Just need the right computer and software behind the cameras.
If Musk had that computer and that software he'd be busy selling the personal assistants from "I, Robot".
Live today, because you never know what tomorrow brings
...is what happens when this tech is on every car? It's all very well to test and develop these things in isolation in California, Nevada or Arizona during bright sunlight.
What about forty or fifty vehicles at a busy intersection, all firing ultrasound, LIDAR and/or microwave in every direction, at night, in the rain? The scope for false positives and false negatives is immense.
Or perhaps the makers will modulate all the output with a unique identifier, perhaps the VIN. So then what happens to your privacy when your identity is being broadcast to all and sundry? Further, what happens when someone spoofs your supposedly-unique output? At high speed and in heavy traffic?
Anyone who thinks that true self-driving vehicles will exist in quantity in the real world beyond specifically-controlled niche use-cases any time soon is delusional.
This sig left unintentionally blank.
You make very well-thought points about the technical stuff, but I disagree with your conclusions. I don't think your pessimism is logical based on the points you've made.
Group of sheep on both sides of the road: not as safe, but probably safe. Lamb on one side, ewe on the other? Very dangerous - the lamb will invariably run to its mother as you approach. Where's the ewe-lamb-running algorithm?
Sure, but to be fair if I had to travel to your area and rent a car, not knowing anything about the behavior of sheep, as a human driver I would be pretty terrible compared to you in that situation. But I doubt you'd say I don't deserve to have a driver's license, because that's a pretty unusual circumstance. I also doubt that even in your area, when you were taking the driving written test, that they had a section on whether you understood the intricacies of sheep behavior.
That's not because your experiences aren't valuable in that situation: it's because there's a difference between a driver and an experienced driver. You get the basic rules of the road going, and statistically you're safe enough to be trusted to be on the road. Young drivers are statistically less safe than experienced ones though. And you will tend to gather experience that relates to the particular location you drive in. Case in point, I'm pretty good at anticipating behavior from drivers in my area, but I've driven in another country and the same cues didn't apply. Still, I didn't get into an accident, being less experienced increases your risk, but it can still be an increased yet acceptable risk.
Where am I going with this? Well, autonomous driving is like an inexperienced driver with a lot of raw talent. It has an advantage over every human driver, including you, in that it will **never** get distracted, or tired. It has a field of vision of 360 degrees at all times, instead of having to choose what it will look at now. It has sensors that can see through objects you can't, so when I'm trying to turn in an area where the parked cars obstruct my view of an incoming car? It doesn't have that problem, it sees the incoming car.
Will it make worse decisions than I would given all the same information? During unexpected hazardous conditions, and assuming I'm not driving after not getting much sleep last night because I had to do another all-nighter at work...probably so. But the fact it has more information than I do and invariable alertness means that even though it would get me into accidents I wouldn't get into if I were driving, it nevertheless will get into less accidents, because it will avoid many of those more common accidents that I would get into if I were driving.
Not only that, but the sheep avoidance algorithm? It's coming. We're talking about the early inexperienced autonomous vehicle, but it's not always going to remain that way. Every time there's an accident with an autonomous vehicles there are logs to be examined. Engineers will pour over them and try to figure out if there's anything that can be done to avoid that type of accident. The problems that occur most often are fixed first, but eventually the system will be good enough that anticipation of animal behavior will be significant. At which point an update gets pushed to your car, and it now knows to slow down and be ready to avoid that road-crossing lamb. Because, unlike humans, when *one* autonomous vehicle gains experience, *all* autonomous vehicles gain that experience. And that's reason to be pretty optimistic.
Warning: Opinions known to be heavily biased.