6 ms·

I wish all the ink spilled on controversy over Tesla would instead be spilled on doing basic safety improvement to the road system that are cheap and proven to

I wish all the ink spilled on controversy over Tesla would instead be spilled on doing basic safety improvement to the road system that are cheap and proven to save many lives. Lets be real here, automated driving or not, having actually save roads helps prevent death and harm in many cases. The hyper focus on high tech software by all the agencies engaged in 'automotive security' is totally wrongly focused. What they should actually do is point out how insanely unsafe and broken the infrastructure is, specially for people outside of cars. See: "Confessions of a Recovering Engineer: Transportation for a Strong Town"

2h 4y ago
I disagree. If someone is a bad driver, and causing crashes, we don't say "we need to improve the road system". We suspend the drivers license until they can prove they are capable of being a safe driver. We should hold this software to the same standard. Until it can demonstrate safety at or above human level, it should be outlawed. Road systems should always be worked on, but when a crash happens its usually the drivers fault, except in a minority of cases where bad road engineering is to blame. this self driving is fucking up enough that it cannot be blamed on the roads anymore, if it ever could.
- jjcon 4y ago
  Tesla publishes their safety data quarterly no need for your many assumptions and speculations - Teslas are already much safer than the average driver, especially when autopilot is on https://www.tesla.com/VehicleSafetyReport https://www.tesla.com/VehicleSafetyReport
  - dmix 4y ago
    One interesting bit from a different 3rd party study: > Cambridge Mobile Telematics also found that people driving Teslas were 21% less likely to engage in distracted driving with their phone in their Tesla compared to when they drove their other car. Maybe the software integration helps avoid this? Lots of other cars have much more complicated interfaces to hook up calls and reading texts. My mom struggled to figure out her car even supported Android Auto. It might just be it's a higher end car, but they didn't see it for an EV Porsche > These findings include an analysis of Tesla drivers who also operate another vehicle. These drivers are nearly 50% less likely to crash while driving their Tesla than any other vehicle they operate. We conducted the same analysis on individuals who operate a Porsche and another vehicle. In this case, we observed the opposite effect. Porsche drivers are 55% more likely to crash while driving their Porsche compared to their other vehicle. The reduction in speed is likely influenced by automated driving, especially considering how fast a Tesla car can accelerate vs normal cars: > They were 9% less likely to drive above the speed limit. https://electrek.co/2022/05/27/tesla-owners-less-likely-crash-than-their-other-cars/ https://electrek.co/2022/05/27/tesla-owners-less-likely-cras...
    - normaljoe 4y ago
      > Maybe the software integration helps avoid this? With auto pilot on the car is watching you. If you take eyes off the road it issues more pay attention nags. Failure to comply removes FSD Beta. So you have a feedback loop where paying attention becomes more important then your phone.
  - super_flanker 4y ago
    Autopilot is not FSD, this report is clearly about Autopilot, they haven't mentioned anything about FSD in it.
    - jjcon 4y ago
      FSD is an extension of the autopilot capability and the autopilot capability has proven extremely safe. As soon as there is statistically enough data they will release FSD numbers too but that means tens of millions of miles logged on roads since crashes are generally uncommon
  - janalsncm 4y ago
    Is autopilot available for all roads? If not, that’s a significant and unstated statistical bias in this data. To have a true comparison, you will need to include only human accidents on roads where Teslas can use autopilot.
- sacrosancty 4y ago
  [dead]
- panick21_ 4y ago
  > I disagree. If someone is a bad driver, and causing crashes, we don't say "we need to improve the road system". Well yes, and that is literally exactly the problem. That is exactly why its so unsafe in the US. Because instead of building a safe system everything is blamed on people. In countries that take road safety seriously, every crash is analyzed and often the road system is changed to make sure it does not happen again. That is why places like Finland, Netherlands and so on have been consistently improving in terms of death and harm caused by the road system. Again, the book I linked goes into a lot of detail about road safety engineering. > We suspend the drivers license until they can prove they are capable of being a safe driver. An unsafe designed street often leads to situation where even good drivers intuitively do the wrong thing. Again, this is exactly the problem. If you build a system where lots of avg. drivers make accidents, then you have a lot of accidents. > We should hold this software to the same standard. Until it can demonstrate safety at or above human level, it should be outlawed. Yes, but its a question of how much limited resources should be invested in analyzing and validating each piece of software by each manufacturer. In general software like Tesla AP would likely pass this test. I am not against such tests but the reality is that resources are limited. > Road systems should always be worked on, but when a crash happens its usually the drivers fault, except in a minority of cases where bad road engineering is to blame. I strongly disagree with this statement. Its a totally false analysis. If a system is designed in a way known to be non-intuitive and leading to a very high rate accidents then its a bad system. Just calling everybody who makes a mistake a bad drive is a terrible, terrible approach to safety. Once you have a safe road system, if somebody is an extremely bad driver, yes taking that person of the road is good. However in a country where so much of the population depends on a car, that punishment can literally destroy a whole family. So just applying it to anybody who makes a mistake isn't viable, specially in system that makes it incredibly easy to make mistakes. The numbers don't even show the problem, the unsafe road system leads to less people walking in the US, and somehow still creating a high rate of deaths for people who walk.
  - eternityforest 4y ago
    America hates any solution to a problem that doesn't involve training individuals to act differently. I constantly hear people who want warning labels gone and periodic PSAs against mixing bleach and ammonia removed, people who don't want mandatory baby left in a hot car detectors, people who hate OSHA, etc. We think the solution should always be human. That the way to solve climate change is homesteads, the solution to poverty is individuals working harder, etc. When anything we don't want happens, we think "I'll try harder next time" not "We will eliminate this as a possibility". People treat life as a sport rather than an engineering challenge, they want to "win fairly", not make it impossible to lose, they want everything they do to say something about their own ability rather than say something about a clever process that makes individual skill irrelevant. I really don't exactly like that aspect of humanity, constantly making everything into a sport.
- ignormies 4y ago
  There's a book called "There Are No Accidents" that goes into depth discussing this fallacy that you should read. https://nyc.streetsblog.org/2022/02/15/excerpt-there-are-no-accidents-yet-we-always-blame-the-victim/ https://nyc.streetsblog.org/2022/02/15/excerpt-there-are-no-...
  - 2h 4y ago
    apparently youve never heard the phrase "texting while driving" or "drunk driving"
    - panick21_ 4y ago
      Its just how incredibly wrong you are. Go look at Salt Lake City. Its is true, Salt Lake City despite horrible bad road design, like Huston. But because people drink less they have slightly less accidents. However international they are still terrible. And that is with there being very little walking. If they had pedestrian and cyclist numbers like Amsterdam it would be 24/7 mass murder.
- janalsncm 4y ago
  At an individual level that makes sense. But at a city level that doesn’t work. People will still need to get around.
orthecreedence 4y ago
If we're actually serious about this, we should get rid of roads altogether and replace them with rail tracks. That solves 95% of the automation problem anyway.
- Zanneth 4y ago
  Replacing roads with railways means nearly complete reliance on a central authority for all transportation needs. This authority decides who gets to leave town and who doesn’t, which companies have priority when delivering goods, how or if the parallel railways should scale to meet more transportation demand, etc. Transportation safety is important but it shouldn’t be considered in isolation when there are potentially catastrophic consequences to prioritizing safety at the cost of everything else.
  - orthecreedence 4y ago
    Why can't personal vehicles use rail tracks without permission from a central authority?
  - achenet 4y ago
    with roads, you already have a central authority (the DMV) that decides who can drive and who can't. Try taking your home-built motorbike with no licence plate out for a spin on public roads, you'll likely get pulled over by police.
- SomaticPirate 4y ago
  US passenger rail is unsafe. We have no way near the level of sophistication of European passenger rail transport. We have a significantly higher number of derailments. Even the worst European rail is more safe than US rail.
  - p_j_w 4y ago
    >US passenger rail is unsafe. US passenger automobiles are more unsafe.
  - 4y ago
marricks 4y ago
Something is changing and it’s very broken so people are paying attention to it. If people really wanted to fix transportation that’s great, high speed rail and public transportation reducing the number of cars on the road seem to be the best solution. But hey, Elon’s hyper loop was a publicity stunt to discourage investment in that. So I say, whether you want to shit on Tesla or public roads, shit on Elon.
- panick21_ 4y ago
  > reducing the number of cars on the road seem to be the best solution. No actually its actually not. Less concession in a system that depends on concession for safety will lead to more accident not less. That is what was shown during Covid, less driving, but accidents per mile went up. So yes, of course public transport, bikes are great, but if you don't fix the underlying problem in the road system, you are gone have a whole lot of accidents. > But hey, Elon’s hyper loop was a publicity stunt to discourage investment in that. This is a claim some guy has made, not the truth. What is more likely is that Musk actually thinks Hyperloop is great (its his idea after all) and would have wanted investment in it. > shit on Elon I prefer not to shit on people most of the time. Musk is the outcome of a South Africa/American way of thought that is more in line with the US avg then most people who advocate for public transport. That is the sad reality. And the problem in the US road system or the US bad public transport can 100% not be blamed on him. There are many people with far more responsibility that deserve to be shit on far more.
  - philosopher1234 4y ago
    What is the underlying problem? I keep looking for it in your comments
    - monissiddiqui 4y ago
      Not GP, but the underlying problem to me is repeat zero-sum interactions in a crony capitalism based system. All the shit stems from the winners that emerge from these repeated games making short term decisions to either allow them to exit or play again.
dzhiurgis 4y ago
I spent 6 months driving a 2019 Hyundai with lane assist and radar cruise control. Personally it's almost perfect. If you added some smart road features - to improve lane assist and better sign readability (to drop the speed as I enter town or curve and increase when I leave one). I don't need full self driving, just an improvement on speed control and some lane assist. Would smart roads be expensive? RFID responders seem super cheap compared to how much actual asphalt costs. Authorities are currently unable to remotely control flows, speeds and safety which is completely bonkers.
- panick21_ 4y ago
  > Would smart roads be expensive? RFID responders seem super cheap compared to how much actual asphalt costs. Authorities are currently unable to remotely control flows, speeds and safety which is completely bonkers. Yeah its not actually that easy. Go and look into train signaling. And cars are not even able to do coupling. Making a train system operate like a super-railway with cars is crazy difficult and has never been done before.
  - int_19h 4y ago
    But is it more crazy difficult than making a car that knows how to safely drive on roads designed specifically for humans? That said, I think even just basic machine-readable road metadata like signs, speed limits, lanes etc would improve matters.
    - panick21_ 4y ago
      > That said, I think even just basic machine-readable road metadata like signs, speed limits, lanes etc would improve matters. That assumes that sign reading is the primary issue with AI cars and it isn't. If you are willing to go to that expense and infrastructure investment, why not just build thing like trains (trams, subways, S-Bahn) and things like Trolley buses and so on. They are far more space efficient, you get much higher threwput then your untested fancy road infrastructure. Another issue with your solution is that there are 1 billion cars out there that do not have any of these things so after decades of working on it, you will still have decades where most cars don't use any of this stuff.
parentheses 4y ago
We’re not talking about road safety. Road safety relies on a baseline amount of driving safety (car hardware + car software). If a car doesn’t stop at a stop sign, no amount of “safe roads” prevent it from mowing down a cyclist. Saying this is like saying we should ignore sex offenders in favor of reworking our society. Instead, we should solve both problems rather than bicker over priorities.
- panick21_ 4y ago
  Ok we should be solving both problems, I agree, but the reality is agencies and political processes have limited capacity. And how much society and influential people in that society is consumed with one thing, the less they do about the other. And I would say that I'm not proposing to rework society in sociological sense, but rather to throw out the standard engineering standards and replace them with better standards. And actually it does matter even in the case you suggest. If a car rolls into an intersection, what speed that car is matters. It matters from what points it is clear that the car is out of control. It matters if there are speed bumps or something along those lines that can send strong signals to a driver. If you have all raised intersection then the top speed of cars will simply be lower and if somebody human or AI makes a mistake that lead to a crash, that crash will be at far lower speed. And proper road design also leads to less intelligence and fancy car design being required. I rather get hit by a shitty designed unsafe old car at 20mph then a fancy new car at 30mph.

2h 4y ago

I disagree. If someone is a bad driver, and causing crashes, we don't say "we need to improve the road system". We suspend the drivers license until they can prove they are capable of being a safe driver. We should hold this software to the same standard. Until it can demonstrate safety at or above human level, it should be outlawed. Road systems should always be worked on, but when a crash happens its usually the drivers fault, except in a minority of cases where bad road engineering is to blame. this self driving is fucking up enough that it cannot be blamed on the roads anymore, if it ever could.

jjcon 4y ago
Tesla publishes their safety data quarterly no need for your many assumptions and speculations - Teslas are already much safer than the average driver, especially when autopilot is on https://www.tesla.com/VehicleSafetyReport https://www.tesla.com/VehicleSafetyReport
- dmix 4y ago
  One interesting bit from a different 3rd party study: > Cambridge Mobile Telematics also found that people driving Teslas were 21% less likely to engage in distracted driving with their phone in their Tesla compared to when they drove their other car. Maybe the software integration helps avoid this? Lots of other cars have much more complicated interfaces to hook up calls and reading texts. My mom struggled to figure out her car even supported Android Auto. It might just be it's a higher end car, but they didn't see it for an EV Porsche > These findings include an analysis of Tesla drivers who also operate another vehicle. These drivers are nearly 50% less likely to crash while driving their Tesla than any other vehicle they operate. We conducted the same analysis on individuals who operate a Porsche and another vehicle. In this case, we observed the opposite effect. Porsche drivers are 55% more likely to crash while driving their Porsche compared to their other vehicle. The reduction in speed is likely influenced by automated driving, especially considering how fast a Tesla car can accelerate vs normal cars: > They were 9% less likely to drive above the speed limit. https://electrek.co/2022/05/27/tesla-owners-less-likely-crash-than-their-other-cars/ https://electrek.co/2022/05/27/tesla-owners-less-likely-cras...
  - normaljoe 4y ago
    > Maybe the software integration helps avoid this? With auto pilot on the car is watching you. If you take eyes off the road it issues more pay attention nags. Failure to comply removes FSD Beta. So you have a feedback loop where paying attention becomes more important then your phone.
- super_flanker 4y ago
  Autopilot is not FSD, this report is clearly about Autopilot, they haven't mentioned anything about FSD in it.
  - jjcon 4y ago
    FSD is an extension of the autopilot capability and the autopilot capability has proven extremely safe. As soon as there is statistically enough data they will release FSD numbers too but that means tens of millions of miles logged on roads since crashes are generally uncommon
- janalsncm 4y ago
  Is autopilot available for all roads? If not, that’s a significant and unstated statistical bias in this data. To have a true comparison, you will need to include only human accidents on roads where Teslas can use autopilot.
sacrosancty 4y ago
[dead]
panick21_ 4y ago
> I disagree. If someone is a bad driver, and causing crashes, we don't say "we need to improve the road system". Well yes, and that is literally exactly the problem. That is exactly why its so unsafe in the US. Because instead of building a safe system everything is blamed on people. In countries that take road safety seriously, every crash is analyzed and often the road system is changed to make sure it does not happen again. That is why places like Finland, Netherlands and so on have been consistently improving in terms of death and harm caused by the road system. Again, the book I linked goes into a lot of detail about road safety engineering. > We suspend the drivers license until they can prove they are capable of being a safe driver. An unsafe designed street often leads to situation where even good drivers intuitively do the wrong thing. Again, this is exactly the problem. If you build a system where lots of avg. drivers make accidents, then you have a lot of accidents. > We should hold this software to the same standard. Until it can demonstrate safety at or above human level, it should be outlawed. Yes, but its a question of how much limited resources should be invested in analyzing and validating each piece of software by each manufacturer. In general software like Tesla AP would likely pass this test. I am not against such tests but the reality is that resources are limited. > Road systems should always be worked on, but when a crash happens its usually the drivers fault, except in a minority of cases where bad road engineering is to blame. I strongly disagree with this statement. Its a totally false analysis. If a system is designed in a way known to be non-intuitive and leading to a very high rate accidents then its a bad system. Just calling everybody who makes a mistake a bad drive is a terrible, terrible approach to safety. Once you have a safe road system, if somebody is an extremely bad driver, yes taking that person of the road is good. However in a country where so much of the population depends on a car, that punishment can literally destroy a whole family. So just applying it to anybody who makes a mistake isn't viable, specially in system that makes it incredibly easy to make mistakes. The numbers don't even show the problem, the unsafe road system leads to less people walking in the US, and somehow still creating a high rate of deaths for people who walk.
- eternityforest 4y ago
  America hates any solution to a problem that doesn't involve training individuals to act differently. I constantly hear people who want warning labels gone and periodic PSAs against mixing bleach and ammonia removed, people who don't want mandatory baby left in a hot car detectors, people who hate OSHA, etc. We think the solution should always be human. That the way to solve climate change is homesteads, the solution to poverty is individuals working harder, etc. When anything we don't want happens, we think "I'll try harder next time" not "We will eliminate this as a possibility". People treat life as a sport rather than an engineering challenge, they want to "win fairly", not make it impossible to lose, they want everything they do to say something about their own ability rather than say something about a clever process that makes individual skill irrelevant. I really don't exactly like that aspect of humanity, constantly making everything into a sport.
ignormies 4y ago
There's a book called "There Are No Accidents" that goes into depth discussing this fallacy that you should read. https://nyc.streetsblog.org/2022/02/15/excerpt-there-are-no-accidents-yet-we-always-blame-the-victim/ https://nyc.streetsblog.org/2022/02/15/excerpt-there-are-no-...
- 2h 4y ago
  apparently youve never heard the phrase "texting while driving" or "drunk driving"
  - panick21_ 4y ago
    Its just how incredibly wrong you are. Go look at Salt Lake City. Its is true, Salt Lake City despite horrible bad road design, like Huston. But because people drink less they have slightly less accidents. However international they are still terrible. And that is with there being very little walking. If they had pedestrian and cyclist numbers like Amsterdam it would be 24/7 mass murder.
janalsncm 4y ago
At an individual level that makes sense. But at a city level that doesn’t work. People will still need to get around.

orthecreedence 4y ago

If we're actually serious about this, we should get rid of roads altogether and replace them with rail tracks. That solves 95% of the automation problem anyway.

Zanneth 4y ago
Replacing roads with railways means nearly complete reliance on a central authority for all transportation needs. This authority decides who gets to leave town and who doesn’t, which companies have priority when delivering goods, how or if the parallel railways should scale to meet more transportation demand, etc. Transportation safety is important but it shouldn’t be considered in isolation when there are potentially catastrophic consequences to prioritizing safety at the cost of everything else.
- orthecreedence 4y ago
  Why can't personal vehicles use rail tracks without permission from a central authority?
- achenet 4y ago
  with roads, you already have a central authority (the DMV) that decides who can drive and who can't. Try taking your home-built motorbike with no licence plate out for a spin on public roads, you'll likely get pulled over by police.
SomaticPirate 4y ago
US passenger rail is unsafe. We have no way near the level of sophistication of European passenger rail transport. We have a significantly higher number of derailments. Even the worst European rail is more safe than US rail.
- p_j_w 4y ago
  >US passenger rail is unsafe. US passenger automobiles are more unsafe.
- 4y ago

marricks 4y ago

Something is changing and it’s very broken so people are paying attention to it. If people really wanted to fix transportation that’s great, high speed rail and public transportation reducing the number of cars on the road seem to be the best solution. But hey, Elon’s hyper loop was a publicity stunt to discourage investment in that. So I say, whether you want to shit on Tesla or public roads, shit on Elon.

panick21_ 4y ago
> reducing the number of cars on the road seem to be the best solution. No actually its actually not. Less concession in a system that depends on concession for safety will lead to more accident not less. That is what was shown during Covid, less driving, but accidents per mile went up. So yes, of course public transport, bikes are great, but if you don't fix the underlying problem in the road system, you are gone have a whole lot of accidents. > But hey, Elon’s hyper loop was a publicity stunt to discourage investment in that. This is a claim some guy has made, not the truth. What is more likely is that Musk actually thinks Hyperloop is great (its his idea after all) and would have wanted investment in it. > shit on Elon I prefer not to shit on people most of the time. Musk is the outcome of a South Africa/American way of thought that is more in line with the US avg then most people who advocate for public transport. That is the sad reality. And the problem in the US road system or the US bad public transport can 100% not be blamed on him. There are many people with far more responsibility that deserve to be shit on far more.
- philosopher1234 4y ago
  What is the underlying problem? I keep looking for it in your comments
  - monissiddiqui 4y ago
    Not GP, but the underlying problem to me is repeat zero-sum interactions in a crony capitalism based system. All the shit stems from the winners that emerge from these repeated games making short term decisions to either allow them to exit or play again.

dzhiurgis 4y ago

I spent 6 months driving a 2019 Hyundai with lane assist and radar cruise control. Personally it's almost perfect. If you added some smart road features - to improve lane assist and better sign readability (to drop the speed as I enter town or curve and increase when I leave one). I don't need full self driving, just an improvement on speed control and some lane assist. Would smart roads be expensive? RFID responders seem super cheap compared to how much actual asphalt costs. Authorities are currently unable to remotely control flows, speeds and safety which is completely bonkers.

panick21_ 4y ago
> Would smart roads be expensive? RFID responders seem super cheap compared to how much actual asphalt costs. Authorities are currently unable to remotely control flows, speeds and safety which is completely bonkers. Yeah its not actually that easy. Go and look into train signaling. And cars are not even able to do coupling. Making a train system operate like a super-railway with cars is crazy difficult and has never been done before.
- int_19h 4y ago
  But is it more crazy difficult than making a car that knows how to safely drive on roads designed specifically for humans? That said, I think even just basic machine-readable road metadata like signs, speed limits, lanes etc would improve matters.
  - panick21_ 4y ago
    > That said, I think even just basic machine-readable road metadata like signs, speed limits, lanes etc would improve matters. That assumes that sign reading is the primary issue with AI cars and it isn't. If you are willing to go to that expense and infrastructure investment, why not just build thing like trains (trams, subways, S-Bahn) and things like Trolley buses and so on. They are far more space efficient, you get much higher threwput then your untested fancy road infrastructure. Another issue with your solution is that there are 1 billion cars out there that do not have any of these things so after decades of working on it, you will still have decades where most cars don't use any of this stuff.

parentheses 4y ago

We’re not talking about road safety. Road safety relies on a baseline amount of driving safety (car hardware + car software). If a car doesn’t stop at a stop sign, no amount of “safe roads” prevent it from mowing down a cyclist. Saying this is like saying we should ignore sex offenders in favor of reworking our society. Instead, we should solve both problems rather than bicker over priorities.

panick21_ 4y ago
Ok we should be solving both problems, I agree, but the reality is agencies and political processes have limited capacity. And how much society and influential people in that society is consumed with one thing, the less they do about the other. And I would say that I'm not proposing to rework society in sociological sense, but rather to throw out the standard engineering standards and replace them with better standards. And actually it does matter even in the case you suggest. If a car rolls into an intersection, what speed that car is matters. It matters from what points it is clear that the car is out of control. It matters if there are speed bumps or something along those lines that can send strong signals to a driver. If you have all raised intersection then the top speed of cars will simply be lower and if somebody human or AI makes a mistake that lead to a crash, that crash will be at far lower speed. And proper road design also leads to less intelligence and fancy car design being required. I rather get hit by a shitty designed unsafe old car at 20mph then a fancy new car at 30mph.