[lftl]

I can't find the specific comment now, but on previous autonomous driving discussions here, it's been asserted that Tesla (and many other systems) simply can't reliably detect stationary objects at speed. If the systems actually did a naive, "that looks like a stationary object, I'm going to avoid it" they'd barely be able to move in a highway setting.

[simion314]

There are systems that can detect stationary objects, Tesla does not want to use those more reliable systems, so they use the programming methodology of tweaking tweaking the code when a bug appears and hope that was the last one.

Sure some AI gurus will claim that humans can detect this objects only with 2 eyes but they forget to mention that behind those 2 eyes is a brain an actual intelligence that is aware of the world and it's place in the world. So if you can't create an actual intelligence then at least put the effort in feeding correct data into your AI and remove the line of code "if static object ignore because too hard to detect if is a truck or a shadow/paper bag/sign"

[SerLava]

Yeah they want a low unit cost on the camera system, so they are trying to machine learn their way out of the problem. It's not going to work.

A self driving system needs to have a 3d understanding of it's environment. It's not acceptable to crash into paintings by Wile E. Coyote.

[simion314]

From what I understand there is also some code that is there to ignore static objects because maybe are signs or bridges and the AI is too stupid. So you have bad sensors with bad code combo.

[SerLava]

Yeah it's essentially impossible for the system to be good- the limitations force other limitations.

[amelius]

> tweaking the code when a bug appears and hope that was the last one

Tesla should also be prohibited from pushing updates. When the software is updated, the code should be tested for at least X thousand miles and X months before a new version may be pushed. Letting an authority do the pushing avoids the temptation of manufacturers to push updates prematurely. It also solves the problem of unexpected changes for the user at random times.

If safety issues are detected, the automobile (or at least the autopilot) should be disabled until the code has been fixed and fully tested.

[nilstycho]

> There are systems that can detect stationary objects, Tesla does not want to use those more reliable systems

What systems are you referring to? Lidar and radar systems can't reliably detect stationary objects as hazards, because most of the world is stationary and consequently gets filtered out.

[simion314]

Filtering out is the problem that develoeprs/engineers should solve right? I can't accept the idea that the car can hit a wall because "our devs are too incompetent or we are to cheap to use better hardware" , a wall or a truck does not look like a bridge or a shadow or a sign, if the AI is too stupid then is not ready, even if is better then a drunk teen that drives a 15 years old car in some very specific situations.

The lidar/radar detects the objects the AI should label them and decide, if it can't do it then throw more video cards at it and more training or find some competent people that could use more deterministic code to reduce the complexity before the AI attempts to run.

[cptskippy]

The problem is with the sensors, not the AI, and occurs across the industry and isn't exclusive to Tesla. Volvo's Pilot Assist has the same short comings.

https://www.wired.com/story/tesla-autopilot-why-crash-radar/

[simion314]

From my reading of your link you are wrong.

How a code that decides what to ignore or not is a sensor issue? Sensors tell you that there is something in front, sure maybe is a sudden thing and there is no velocity/acceleration vector for a few moments but the object is detected ...the software DECIDES what to ignore.

The example where a car changes slane and the code ignores any static object in front of that car seems a clear case of a crash waiting to happen, the car in front might change the lane to avoid something. (and the fact that some other tech that is honestly labeled lane assist has this issue does not excuse a self claimed autopilot to have same flaws."

Let me summarize and let me know where I am wrong:

"Because the software is bad and it triggers too many false positives the coders added a "clever" hack , we ignore most static objects."

[cptskippy]

> How a code that decides what to ignore or not is a sensor issue?

I guess "Sensor Module" would be a more accurate description.

Most automotive systems are designed as a series of modules that function independently but jointly form a system. They either broadcast or consume messages over the CAN BUS.

The Volvo V60 for example has a Forward Sensor Module, the radar, which outputs a handful parameters like Primary Target Range, Voltage, and Automatic Alignment Offset. It's the responsibility of other body modules to consume that data. The Closing Velocity Module might then broadcast Acceleration related messages which intern would be picked up by the Adaptive Cruise Control Module. Or it might broadcast Collision Warning Messages which the Stereo Module reads and responds by blaring a warning noise out the speakers and also the Brake Control Module receives and applies the brakes.

[nilstycho]

Yes, in principle there are systems that could detect stationary hazards. But as far as I'm aware the best lidar and radar approaches are currently no better than Tesla's RGB approach. It seems that the AI is much more important than the sensor.

[simion314]

The sensors could detect a solid object, the issue seems to be that the sensors will detect many things and the AI or the hardware is incapable to handle the amount of sensors data.

Though the Google car seems to have strong enough hardware to keep track of many vehicles and objects at a big enough distance to not be surprised if something appears or disappears for a frame.

Not sure what people want for a sensor, to detect and label the objects for you? or calculate velocities and predict the future? That seems to be the software developers job but you might need competent people and not throwing big data into a black box and retrain until it mostly works.

[camjohnson26]

No Waymo car has crashed into an emergency vehicle yet.

[nilstycho]

Waymo has driven 20 million miles. Autopilot has driven (about) 4 billion miles.

[cptskippy]

The systems you're referring to are radar based systems. Tesla stopped putting radar in the Model 3/Y in favor of it's vision system.

It would be interesting to see a breakdown of the 12 incidents and the Tesla Models involved. Were the incidents involving radar/vision blended, or vision only vehicles or both?

If the incidents only involved older vehicles that were radar equipped, that would be good PR for the vision only system. If the incidents were in vision only systems, it would be bad. I suspect it'll probably be a little of both.

[Someone]

AFAIK, it can detect them just fine, but it is bad at discriminating those its collision avoidance part should pay attention to from those it shouldn’t pay attention to (e.g. the road surface, traffic signs on the side of or over the road), so they (¿mostly?) chose to ignore all of them.

[sudosysgen]

That just means it lacks the ability to place objects in 3D space and plot a course, which is... sad.

[cptskippy]

Most collision avoidance systems aren't that sophisticated. Radar based systems like those used in Volvo's Pilot Assist suffer from the same inability to discern stationary objects.

Most systems are self contained modules with low power micro-controllers that can emit basic warning and alert data onto the CAN BUS. Tesla's is actually taking the input from it's sensors and attempting to do more in-depth analysis and actually produce a 3D space.

The fact that they're still suffering from the same problem as other manufacturers shouldn't be an admonishment of their technology or attempts unless or until it's proven that their approach is fundamentally worse.

[sudosysgen]

Well, no. No system that is unable to detect and map out a 3D maps of obstacles and plot a course through it without hitting stationary, well visible objects, is not suited to being used in a self-driving capacity. Whether it's from Tesla or anyone else.

If their approach fundamentally means they decided to work around mapping their surroundings in 3D correctly, as the GP suggested, there is a strong reason to admonish.

[MichaelZuo]

That seems like an enormous caveat to the system overall. Do you know if there is any public justification for going along that route, omitting radar, other sensors, etc.?

[ocdtrekkie]

If that's the case, the correct solution is to shut Autopilot and all systems like it down, and charge their developers with reckless endangerment.

A autonomous driving system which can't detect stationary obstacles is not an autonomous driving system: It's a murder machine.

If they can't do this, they shouldn't operate, full stop.

[madamelic]

The horrible truth is that if autonomous driving can save 1+ life, it's better than what we have now so we should keep it and accept that crashes will occur much like they do now with humans in sole control.

There's no reason to kill autonomous driving just because gen 1 - 50 isn't perfect but is an improvement on current driving methodology.

Giving a one-line sentence of what happened disregards other variables such as, lighting conditions, where the stationary object was (Over a hill, round a sharp bend, etc.), how much stopping distance was there. A lot of Tesla hate is purposefully done from EV haters & Musk haters. No one can argue that there isn't an establishment who wants to tear down and prevent Tesla & Musk from succeeding. Teslas will catch on fire (like other cars do), Teslas will crash (like other cars do), Auto-pilot will fail (while also being overall an improvement on driving safety).

We can't assume the stationary object was 300 feet away on a straightaway in a perfect visibility conditions, we have to judge whether a human would've performed worse and whether the same cases are occurring over and over. 12 crashes over a few months is a pretty tiny fraction when there are probably more than 12 humans crashing into stationary objects every hour.

[rutthenut]

I do not agree with this argument

>> The horrible truth is that if autonomous driving can save 1+ life, it's better than what we have now so we should keep it and accept that crashes will occur >>

As that '1 life' saved may be countered by a number of lives lost due to the system that might otherwise have survived.

[hnthrowaway8493]

>if autonomous driving can save 1+ life, it's better than what we have now so we should keep it and accept that crashes will occur

This argument ignores the reality of driving: by being a competent, attentive, careful driver you can avoid many classes of accident. Crashes are all about fault, and as a safe driver you can avoid most at-fault accidents. You get to avoid all the self-inflicted damage of drunk/incapacitated driving, unsafe driving for weather conditions, extreme speeding / reckless driving, etc. You still have to deal with the externalities of those, but some fraction of the time you can notice and anticipate these people on the road and stay away. Not always of course, so that still leaves some number of deaths per mile, but I'd bet that threshold is way lower than the "save one life" bar. So if you're a good driver, autonomous cars at the "current death rate" safety bar would make you less safe.

What you get by ignoring this and using the current death rate as a bar is a "transfer of deaths": you take all the deaths that were the fault of the driver (driver dies due to being drunk/high, very reckless driving, etc) and transfer them to random innocent people who were not at fault.

[vmladenov]

> This argument ignores the reality of driving: by being a competent, attentive, careful driver you can avoid many classes of accident.

It’s funny you use the word reality since this is completely at odds with driving in America. You’re comparing an idealized world of competent drivers to the real world where there are basically no barriers to getting a license and everyone is on their phone behind the wheel. Unless you advocate for stricter licensing and more police enforcement against inattentive driving with equal vigor, this line of reasoning doesn’t hold.

[_jal]

>can save 1+ life, it's better than what we have now so we should keep it and accept

So I take it you would be fine with the nightmare version of Apple's phone scanner, continuously scanning your devices for thoughtcrime? After all, "If it saves just one child..."

Any time I hear emotional appeals that ignore the cost side of the equation, I look for a financial motive.

[Dylan16807]

Let's rephrase. If the important upsides and downsides are all represented in the number of traffic fatalities, and that number goes down, then that's good.

The problem with the Apple system is significant downsides that are completely separate from CSAM.

[a_square_peg]

“The horrible truth is that if autonomous driving can save 1+ life, it's better than what we have now”

This is not true at all and we have much higher standards than this. This is like saying if a vaccine cures Covid but manages to kill people at a rate slightly less than Covid, we should embrace it.

[schmidp]

Isn’t that exactly what we are doing? (-slightly, as we would need to define the threshold)