First Tesla Autopilot Death

Seen something interesting in the news or on the intertubes? Discuss it here.

Moderators: Zamfir, Hawknc, Moderators General, Prelates

morriswalters
Posts: 7073
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: First Tesla Autopilot Death

Postby morriswalters » Mon Jul 04, 2016 11:45 am UTC

BattleMoose wrote:I was hoping to learn what happened. Why the did what they did and why it failed.
Perhaps they were over ambitious. But a successful downtown is more than pedestrian spaces. Today much of it has been partially opened to traffic and it is doing better as suburban areas are static. But if I had to guess I'd say it was the lack of people after working hours, a situation that is changing today.
commodorejohn wrote:I know that they're just one of many places in which J. Random Troublemaker can find huge numbers of cars relatively unattended!
Relatively? I suspect if you set up converting 20 or so cars to bombs one of those slack jawed security people might notice, not to mention the relatively brisk turnover in cars by their owners. This particular idea fails on my stinkum meter.
commodorejohn wrote:Also, can I just say that I love how every argument for self-driving cars eventually winds up back at the most basic defense of "well, at least they haven't killed as many people as human drivers! (Yet!)" It's wonderful.
Since there are so few self driving cars what else can you say. The market will determine the validity of the concept

RCT Bob
Posts: 75
Joined: Sun Aug 02, 2015 2:05 pm UTC
Location: Netherlands

Re: First Tesla Autopilot Death

Postby RCT Bob » Mon Jul 04, 2016 11:49 am UTC

BattleMoose wrote:
Everybody should ride bikes, to the best of my knowledge they are the only common form of transport that weighs less than.


Everybody should! The benefits are outrageous!

People would be healthier and happier. Reduced costs medical costs. Won't need nearly as many parking spots in prime locations. Air quality would be better. There would be so much less noise. Not to mention climate change. And the maintenance on roads would be so much less. And people would be less stressed too. Have you noticed how stressful driving can be and how angry it makes some people?

It's actually just so nice when there aren't cars around. The problem is to get these benefits you need an entire societal shift towards bicycles. In the mean time there are too many meatsacks driving too recklessly to make cycling pretty damn dangerous. To make cycling delightful, you have to get rid of the cars but you can't do that until cycling is delightful....

(I don't own a car, its not that hard. Cars are stupid. Rather private ownership of cars is stupid.)


You should come to the Netherlands, everyone here has a bike. But bikes also require parking space

Image
Better remember where you parked your bike
Image

Image
Yes, in some places we even park them on separate floors

On most roads we actually have separate bicycle lanes, so there is the main road for cars, the bicycle lane for cyclists (and people on skateboards etc.) and the sidewalk for people on foot. Most shorter distances (up to 10-15km) are travelled by bicycle in the Netherlands. Granted, fuel prices are also significantly higher in Europe compared to the US (around EUR 1.50/L for unleaded fuel).

Image

User avatar
Dauric
Posts: 3886
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: First Tesla Autopilot Death

Postby Dauric » Mon Jul 04, 2016 1:45 pm UTC

Another factor with cycling is the distances involved. In Colorado (and most western U.S. states) commutes of 60 miles are fairly commonplace, and public transit is often... lacking. If you live in a fairly urban area to begin with then it makes sense, but in semi-rural and rural areas (which have tax-bases too low to support an effective mass transit system) there's not a lot of options other than individual automobile ownership.
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

BattleMoose
Posts: 1993
Joined: Tue Nov 13, 2007 8:42 am UTC

Re: First Tesla Autopilot Death

Postby BattleMoose » Mon Jul 04, 2016 1:51 pm UTC

Dauric wrote:Another factor with cycling is the distances involved. In Colorado (and most western U.S. states) commutes of 60 miles are fairly commonplace, and public transit is often... lacking. If you live in a fairly urban area to begin with then it makes sense, but in semi-rural and rural areas (which have tax-bases too low to support an effective mass transit system) there's not a lot of options other than individual automobile ownership.


Truly. But think of the potential for all our very high density cities, especially CBDs.

User avatar
Dauric
Posts: 3886
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: First Tesla Autopilot Death

Postby Dauric » Mon Jul 04, 2016 2:08 pm UTC

BattleMoose wrote:Truly. But think of the potential for all our very high density cities, especially CBDs.


Sure like I said, in an already urbanized environment cycles and auto-taxis make sense. They're not a universal panacea to the issues of personal transit however, and consideration does have to be given to how multiple systems interact.

Where I live I can see as effective a mix of personally owned hybrid manual/autonomous vehicles (total one way or the other, not the driver-supposed-to-be-aware-but-probably-isn'the that the Tesla car in this incident had). The tax base is too spread out to support high concentrations of mass transit, even autonomous taxis, and once you get about 60 to 80 miles out of Denver you start hitting rural areas where you routinely encounter dirt roads and poor road maintenance, so the visual cues your autonomous cars need frequently dissappear.

You want to avoid the situation where you've got such specialized solutions for your urban, suburban and rural environments that transitioning between those areas creates difficulties that incentivises (ironically) a lack of mobility.
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7427
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: First Tesla Autopilot Death

Postby Zamfir » Mon Jul 04, 2016 2:29 pm UTC

Yeah, that's already a serious issue with bikes, even in the Dutch cycling paradise. Cycling to work requires either that you live close to many potential employers (usually the most expensive housing), or you make proximity to your house the overriding factor for work, or you move to one specific employer, which makes you vulnerable. Especially with the constant trend towards more 'flexible' employment.

About 30% of people here take their bike to work, which takes up less than 10% of everyone's combined commuting kilometers. The catch is obvious: cycling is most attractive, if you had an easy commute anyway.

BattleMoose
Posts: 1993
Joined: Tue Nov 13, 2007 8:42 am UTC

Re: First Tesla Autopilot Death

Postby BattleMoose » Mon Jul 04, 2016 2:56 pm UTC

I was a very keen cyclist, and my commute was 12km, which was absolutely fine for me. I take public transport now, which takes longer (much longer if connections are bad), costs more money, doesn't count as exercise and you are restricted to when and how it decides to operate. The distance to me just wasn't an issue, would happily commute a longer distance.

I really cannot over emphasize how unimportant the distance aspect was for me. The important issues are how pleasant the route is and how safe it is. And if it is pleasant and safe, the commute is just, great. Its an enjoyable life experience and should it take longer, while, I get to enjoy this experience for longer. I think at 20km it starts to be looking like a bit far and in a large city, that covers an enormous radius. I appreciate not everyone shares this perspective but there are many who do and if it could be culturally normalized I think it could become very popular.

EDIT: And I have been in and out of many gyms so far in my life. And always the most popular piece of equipment is the treadmill or some other cardio piece of equipment. They are unreasonably popular even in homes. Just turn your commute into your cardio and you don't have go to gym and then you can save all this time and money. Its a just a better way.

Spoiler:
I was hit by car a bit over a year ago, it was a very low speed accident and while I escaped uninjured it rattled me up a very great deal. Had so many near misses too. Being still reasonably young the probability of encountering an truly awful accident at some point in my life was just too real for me.

Its also worth emphasizing how absolutely awful the attitude here in Australia is towards cyclist. Viewed as vermin but the majority of the public. Hit and runs are reasonably common, usually by accident, sometimes not. We even have some idiot who regularly spreads nails and tacks along a popular cycle route, because hate.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3369
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: First Tesla Autopilot Death

Postby Soupspoon » Mon Jul 04, 2016 3:23 pm UTC

Zamfir wrote:cycling is most attractive, if you had an easy commute anyway.
And if it's flat. Can't predict the wind and other weather, day to day, but you at least know your journeys are flat. (Several decades ago, when I did a cycle-tour across the area, we were lucky that we had a tail-wind strongly gusting us across the Afsluitdijk... If anyone actually cycle-commutes that 20 miles, the opposite journey would have been soul-destroying on that day.)

I live in a somewhat hilly area. It's not so bad, but a bit discouraging. Even if we had decent bike-friendly infrastructure (by which I mean more secure bike-parking in retail areas, etc, I'm not a fan of highway apartheid), people who aren't already fit wouldn't be keen to cycle a mile or two rather than take the car/etc. Last weekend there was a competitive event up a hill, closer to my way, that some people will avoid walking up, and would even take the bus...


@Battlemoose: I recognise (a variation of) your experiences, even from t'other side of the planet. We're not so much seen as vermin, but it doesn't help that the popular opinion (especially in a city that has put shared cycle/pedestrian paths across the area) is that a bicycle is not a proper road vehicle. Which is the inverse of the legal reality. My heart sinks when an adult1 is riding on a (non-permissive) pavement and perpetuating fellow car-drivers' (and of larger vehicles) misconceptions that they have exclusive use of the road. More cycles on the road (obeying all lights, etc, of course), please, and please scrap those useless 20-yard off-road stretches of coloured tarmac that serve no useful purpose (except to claim 'green' points for jystifying a claim of so-many miles of cyclepaths existing) and have to swerve around the reverse of pedestrian crossing waiting spots...

1 There's not an official age-cutoff, but I reluctantly hold my ire for youngsters.

User avatar
HES
Posts: 4861
Joined: Fri May 10, 2013 7:13 pm UTC
Location: England

Re: First Tesla Autopilot Death

Postby HES » Mon Jul 04, 2016 3:54 pm UTC

Soupspoon wrote:please scrap those useless 20-yard off-road stretches of coloured tarmac that serve no useful purpose (except to claim 'green' points for justifying a claim of so-many miles of cyclepaths existing) and have to swerve around the reverse of pedestrian crossing waiting spots...

As someone who designs cycle infrastructure for a living, nothing annoys me more than bad cycle infrastructure. Not only is it useless, it's counter-productive. It is getting better, though.

As a general response to several other posts - Most barriers to cycling are actually fairly easy to overcome. Having a city laid out so that people live 60 miles from work, is not one of them.
He/Him/His Image

User avatar
Dauric
Posts: 3886
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: First Tesla Autopilot Death

Postby Dauric » Mon Jul 04, 2016 4:11 pm UTC

HES wrote:As a general response to several other posts - Most barriers to cycling are actually fairly easy to overcome. Having a city laid out so that people live 60 miles from work, is not one of them.


In Colorado at any rate it's not so much that "a city" is laid out with 60-mile commutes, but that the major metropolitan region is that spread out. For instance I work at a financial institution in the downtown Denver area, but I live 60 miles east in the city of Aurora, housing just isn't that cheap in Denver itself.

You're not looking at a cohesive "urban plan", it's a stretch of multiple (theoretically) independent municipalities with their own elected city governments. It's called "Urban Sprawl".
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

User avatar
ucim
Posts: 6354
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: First Tesla Autopilot Death

Postby ucim » Mon Jul 04, 2016 6:32 pm UTC

HES wrote:Most barriers to cycling are actually fairly easy to overcome. Having a city laid out so that people live 60 miles from work, is not one of them.
Neither is weather. Neither is terrain. Neither are incompatible roads. It's gotta be built in from the getgo. California, for example, has a new law that one must give bicyclists three feet clearance when passing them. Good idea, except that roads simply don't have the space to do that (without breaking other laws, like crossing the double yellow line, which is (expressly, I believe) forbidden by the bicycle law. The upshot is that one bicycle forces legal car traffic to bicycle speed, until the cyclist can get off the road (and decides to do so) and let the cars pass.

I wonder how different the reaction to the Tesla crash would have been had he killed a cyclist instead of himself while "driving" on autopilot.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
HES
Posts: 4861
Joined: Fri May 10, 2013 7:13 pm UTC
Location: England

Re: First Tesla Autopilot Death

Postby HES » Mon Jul 04, 2016 6:57 pm UTC

He/Him/His Image

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7427
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: First Tesla Autopilot Death

Postby Zamfir » Mon Jul 04, 2016 8:23 pm UTC

Neither are incompatible roads. It's gotta be built in from the getgo

Weather and terrain, sure. But you have to convince me about this. Most bike infrastructure around me has been put in place later on, it always seems to fit if they try hard enough. Cyclists takes up less space than cars, it's a space saver if you can switch traffic from one mode to the other.

Can you post some pictures (or street view links) of roads that are incompatible?

User avatar
Dauric
Posts: 3886
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: First Tesla Autopilot Death

Postby Dauric » Mon Jul 04, 2016 8:45 pm UTC

Zamfir wrote:
Neither are incompatible roads. It's gotta be built in from the getgo

Weather and terrain, sure. But you have to convince me about this. Most bike infrastructure around me has been put in place later on, it always seems to fit if they try hard enough. Cyclists takes up less space than cars, it's a space saver if you can switch traffic from one mode to the other.

Can you post some pictures (or street view links) of roads that are incompatible?


I believe the situation Jose is talking about is it's not so much that they put in actual infrastructure that doesn'the work, they just legislated rules of the road without going to the expense of actually installing actual infrastructure (or other flavors of essentially "not trying hard enough"). It's kind of like that useless segments of bike path thing, measures to make some politicians look good to the Green voters without actually incurring the expense a proper bike infrastructure would require. You end up with a hodge-podge of dedicated bike paths, and segments of unmodified automobile road "connecting" them together.
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

User avatar
LaserGuy
Posts: 4516
Joined: Thu Jan 15, 2009 5:33 pm UTC

Re: First Tesla Autopilot Death

Postby LaserGuy » Mon Jul 04, 2016 9:36 pm UTC

Incidentally, here's a story of a bus running a red light and t-boning a tractor trailer, resulting in 4 deaths and 25 injured.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3369
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: First Tesla Autopilot Death

Postby Soupspoon » Mon Jul 04, 2016 9:49 pm UTC

Zamfir wrote:Can you post some pictures (or street view links) of roads that are incompatible?

That'd be my question too. I also have my own answer (that when roads get too optimised for high-speed traffic, they get given incompatabilities with cycling, rather than just lack designed-in compatibility features1) but a basic, reasonably well-maintained road should not be considered out-of-bounds to any road vehicle use.

Motorway restrictions (~2,000 miles out of ~245,000 miles, in the UK) obviously disbar cyclists, and I wouldn't ride on them. And there are very select stretches of major A-roads with specific restrictions applied, but I'd be more than willing to use a majority of the the 30,000 miles of A-roads. Aside from th swathes of 'near-Motorway' design used in major trunk A-roads, the traditional main road and the >200,000 miles of B, C and U-roads suffer not from bad road design, but driver behaviour.

1 Which is not to say that features can't still be added, but directing cyclists to mount the footway shortly before every junction, instruct them to dismount to cross each side-road and then bringing them back onto the road afterwards 'lacks something', when it's far easier (if you can trust the other traffic to do as they are supposed to) to just ride along the road (or on the green/red-coloured metre-strip, if necessary) with all the right of way of a normal vehicle. Additional segregation doesn't help drivers appreciate that bicycles, horses, etc are as legitimate users of the road as they, for all the hundreds of thousands of miles where there is no 'special provision'. (I hope I'm in tune with you on this, HES. Obviously I've been 'on the road' , man and boy (and bike and car and pedestrian and occasionally even minibus or white van) for a number of decades, and I've seen (for example) the A1 turn from safe enough to ride on (but still busy, relative to other roads) to something only marginally short of a two-lane motorway (especially including the now 'seemless' transition from A1(M) to A1, at various points, where the major traffic island junction has been transformed into a straight-through-with-sliproads-and-overpass junction that means drivers may not even realise that they're not now under Motorway rules and conditions) that has seen a number of cycle fatalities from SMIDSY drivers of car and lorries alike. (That's a "cyclists designed out" road, if I ever saw one...)


(I'm talking from a UK perspective here, but apart from Motorways, specifically, and individually regulated special cases, there's no such thing as an 'automobile road'. Indeed, the widest definition of 'road' or 'highway' includes footpaths, bridleways and drovers' trails, down which the more modern motor vehicle has no, or limited, permission to travel. But I'll happily just refer to all metalled carriageways publicly (or toll-wise) maintined, for simplicity.)

User avatar
Dauric
Posts: 3886
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: First Tesla Autopilot Death

Postby Dauric » Tue Jul 05, 2016 12:14 am UTC

In the U.S. the existence of pedestrian walkways (sidewalks) isn't consistent. Even in major cities you can get stretches of no pedestrian walkways, and they're practically unheard of in rural areas.

Even in areas of consistent walkways, bicycles may not be permitted to be ridden on walkways. In a typical case of shitheads ruining it for everyone, enough accidents between cyclists and pedestrians and you get regulations where cyclists (and skateboards, and wheeled conveyances of any type) can get ticketed for cycling on walkways.
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

User avatar
ucim
Posts: 6354
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: First Tesla Autopilot Death

Postby ucim » Tue Jul 05, 2016 1:06 am UTC

Zamfir wrote:Can you post some pictures (or street view links) of roads that are incompatible?
Pretty much any rural two lane two way road with a double yellow line, on which cars drive at natural car speeds. Many of the roads connecting towns in California are like that. Sure, you can ride on them, and cars can drive on them, but a car at 55 mph passing a bicyclist at 10 mph is going to be a problem when there's no legal place for the car to go to give the bike a safe buffer.

Side streets aren't a problem really. Despite speeding cars, they can cross the center to pass a cyclist. But inter-town roads (which often are major arteries in more rural areas) often have one lane each direction, a double yellow line between, curves and hills, cars (legally) doing highway speed, and no place to widen the road to add a proper bike lane without major construction.

Here's an example pulled more or less at random.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

chris857
Posts: 100
Joined: Thu Dec 02, 2010 5:04 pm UTC

Re: First Tesla Autopilot Death

Postby chris857 » Tue Jul 05, 2016 6:16 am UTC

Soupspoon wrote:
Zamfir wrote:I live in a somewhat hilly area.


If I had to bike to work, I'd have to go uphill about 400 vertical feet each way (there's a valley between home and work). If it were flat, I might be able to handle the 8.5 miles each way, but when even one 1/2 mile section defeats me with the steepness, it kind of puts it out of the question.

Tyndmyr
Posts: 11028
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: First Tesla Autopilot Death

Postby Tyndmyr » Tue Jul 05, 2016 2:35 pm UTC

ahammel wrote:
Liri wrote:It's amusing to imagine millions of driverless cars shuttling people back and forth from their homes to work instead of using some form of public transportation.

The cars might well be the public transportation.


This, really.

There's no particular reason that public transportation HAS to mean packed in a large vehicle with a bunch of strangers. There's a number of downsides to that, it's just done because it's efficient. Kind of. For certain population densities. DC sucks at it badly.

For some densities, car is the only thing that makes sense, and if you do away with inconveniences like drivers and parking, cars become significantly better in high density areas as well.

If you have a driver that never gets tired, running a car 24/7(or whatever a reasonable maint cycle allows) seems likely to be very efficient indeed. It might be owned by a company or individuals instead of cities, but regardless of the money back end, it's essentially a public transport system at that point. Potentially emergent from an uber-like system displacing everything else gradually.

Kind of nice to think about.

User avatar
sardia
Posts: 6254
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: First Tesla Autopilot Death

Postby sardia » Tue Jul 05, 2016 2:56 pm UTC

Autonomous driving will probably be realized for large vehicles first. Given the cost of the sensor package, it makes more sense to spend thousands of dollars for one large bus or tractor trailer compared to a relatively cheap car. More importantly, they follow set routes, and don't have to be programmed for all the crazy spots people want to drive to.

User avatar
ucim
Posts: 6354
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: First Tesla Autopilot Death

Postby ucim » Tue Jul 05, 2016 3:33 pm UTC

sardia wrote:Autonomous driving will probably be realized for large vehicles first.
You mean, the ones where a crash is really bad? Yep, that's where I'd put the experimental unproven tech</sarcasm>.
sardia wrote:More importantly, they follow set routes, and don't have to be programmed for all the crazy spots people want to drive to.
...but that's not the hard part, not by a long shot.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

KnightExemplar
Posts: 5494
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: First Tesla Autopilot Death

Postby KnightExemplar » Tue Jul 05, 2016 4:01 pm UTC

sardia wrote:More importantly, they follow set routes, and don't have to be programmed for all the crazy spots people want to drive to.


You mean like the autonomous trains we already have?

Even in America?
First Strike +1/+1 and Indestructible.

User avatar
HES
Posts: 4861
Joined: Fri May 10, 2013 7:13 pm UTC
Location: England

Re: First Tesla Autopilot Death

Postby HES » Tue Jul 05, 2016 4:08 pm UTC

The most serious collisions are when a HGV veers sideways across three lanes, through the central barrier, and across another three lanes of oncoming traffic. Most commonly attributed to tired, distracted, or unwell drivers. Taking that risk away sounds like a great idea to me.

ucim wrote:Yep, that's where I'd put the experimental unproven tech</sarcasm>.

Tough shit, they'll be on the road within a year.

Being able to stick to controlled highways/motorways between depots makes things significantly easier. No pedestrians and no intersections - if those aren't the "hard part" I don't know what is. After all, autopilot seems to work just fine when there no conflict points.
He/Him/His Image

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3369
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: First Tesla Autopilot Death

Postby Soupspoon » Tue Jul 05, 2016 4:30 pm UTC

HES wrote:The most serious collisions are when a HGV veers sideways across three lanes, through the central barrier, and across another three lanes of oncoming traffic.
Most serious, maybe, but not most common. Cost/benefit analysis might rebalance things towards commoner, less individually 'impactful' (NPI) accidents prevented.

Especially with an unknown rate of system failure maybe being better and more conservatively applied to something other than the big-rig (or, perhaps worse, the free-driving autonomous bus of schoolkids, carrying its own ready-made casualty list, rather than having to go out and find one on the road, to fail near) until a bit more proven to not be more liable to systemically jammed accelerators or somesuch.

cphite
Posts: 1253
Joined: Wed Mar 30, 2011 5:27 pm UTC

Re: First Tesla Autopilot Death

Postby cphite » Tue Jul 05, 2016 4:51 pm UTC

HES wrote:The most serious collisions are when a HGV veers sideways across three lanes, through the central barrier, and across another three lanes of oncoming traffic. Most commonly attributed to tired, distracted, or unwell drivers. Taking that risk away sounds like a great idea to me.


Sure; if you're actually taking that risk away. Not everybody is convinced that it's being taken away.

We already have an example here of an autonomous vehicle failing to notice a tractor trailer. It's not unreasonable to imagine a scenario where one of these vehicles fails to turn at highway speeds, or turns when it should not. Either could happen as a result of sensor malfunction or a more general system malfunction, and give you your sideways across lanes scenario.

ucim wrote:Yep, that's where I'd put the experimental unproven tech</sarcasm>.

Tough shit, they'll be on the road within a year.


Ah, yes... the "tough shit we're doing it anyway!" argument... always a classic.

Being able to stick to controlled highways/motorways between depots makes things significantly easier. No pedestrians and no intersections - if those aren't the "hard part" I don't know what is. After all, autopilot seems to work just fine when there no conflict points.


Pedestrians and other people (and drivers) is one of the hard parts. Another hard part is that we aren't anywhere close to having perfect sensors that work all the time, or perfect computers that work all the time. For things like trains and planes that undergo constant maintenance and monitoring of systems, you can mitigate that... I'm not sure your average private citizen is going to do that.

One thing that a lot of people forget when they're touting the safety record of these vehicles, is that so far the vast majority of data has come from tightly controlled testing units. They're vehicles that are constantly being watched and updated if necessary. We have very little data about how these systems will work when they're sitting in Joe Blow's garage for a year - especially if Joe isn't too keen on upkeep and repair. How well do they handle dirt, damage, missed updates, normal wear and tear, etc?

Sorry, but until we have some actual real world data (that is, data that isn't coming from controlled test conducted by people with a financial interest in the results) it's just too early to assume that these things are safer than human drivers. People can parrot the accidents per miles numbers until the cows come home; those aren't independent tests, nor are they taking into consideration the enormous numbers of variables introduced by the human owner of the vehicle.

User avatar
ucim
Posts: 6354
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: First Tesla Autopilot Death

Postby ucim » Tue Jul 05, 2016 4:54 pm UTC

Trains are easy - they're on rails. And yes, starting with the "easy routes" makes sense, it's just that programming the route isn't the hard part; the hard part is recognizing unexpected intrusions and avoiding colliding with them. Distinguishing between a pedestrian and a paper bag. Distinguishing between a truck and the sky. But sure, on routes where such things are less common, collisions will also be less common.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

commodorejohn
Posts: 1083
Joined: Thu Dec 10, 2009 6:21 pm UTC
Location: Placerville, CA
Contact:

Re: First Tesla Autopilot Death

Postby commodorejohn » Tue Jul 05, 2016 4:56 pm UTC

cphite wrote:
ucim wrote:Yep, that's where I'd put the experimental unproven tech</sarcasm>.

Tough shit, they'll be on the road within a year.

Ah, yes... the "tough shit we're doing it anyway!" argument... always a classic.

It's good to know they're so concerned for the well-being of others.
"'Legacy code' often differs from its suggested alternative by actually working and scaling."
- Bjarne Stroustrup
www.commodorejohn.com - in case you were wondering, which you probably weren't.

User avatar
sardia
Posts: 6254
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: First Tesla Autopilot Death

Postby sardia » Tue Jul 05, 2016 4:57 pm UTC

Cphite, how is that any different than when Joe scmoe misses crucial maintenance and causes deaths anyway in a regular car? Should hybrids be banned because they use complex braking systems?
Last edited by sardia on Tue Jul 05, 2016 5:26 pm UTC, edited 1 time in total.

User avatar
HES
Posts: 4861
Joined: Fri May 10, 2013 7:13 pm UTC
Location: England

Re: First Tesla Autopilot Death

Postby HES » Tue Jul 05, 2016 5:17 pm UTC

cphite wrote:Another hard part is that we aren't anywhere close to having perfect sensors that work all the time, or perfect computers that work all the time.

Nor do we have perfect humans that work all the time.

And this is the fundamental point when it comes to driverless technology.

People will continue to hold machines to vastly higher standards than people - and this is a good thing, because it will push the industry as close as possible to that perfect, unbreakable system. Just so long as we realise that perfect, unbreakable system doesn't exist. If we can't have "best", can we at least have "better"?
He/Him/His Image

speising
Posts: 2252
Joined: Mon Sep 03, 2012 4:54 pm UTC
Location: wien

Re: First Tesla Autopilot Death

Postby speising » Tue Jul 05, 2016 5:23 pm UTC

I'm afraid that we can't. The failure modes of the systems will always be different than the humans'. So that everytime an autonomous vehicle produces an accident people will point out how stupid the systems are and how much better *they* would have handled the situation.
Totally ignoring all the situations where the machines were better…

KnightExemplar
Posts: 5494
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: First Tesla Autopilot Death

Postby KnightExemplar » Tue Jul 05, 2016 5:27 pm UTC

speising wrote:I'm afraid that we can't. The failure modes of the systems will always be different than the humans'. So that everytime an autonomous vehicle produces an accident people will point out how stupid the systems are and how much better *they* would have handled the situation.
Totally ignoring all the situations where the machines were better…


That's not even this case however.

The Tesla "Autopilot" isn't an autonomous system. The driver just believed it was and was watching Harry Potter, not understanding the limitations of the system. Probably because Tesla's marketing implied otherwise.
First Strike +1/+1 and Indestructible.

Chen
Posts: 5440
Joined: Fri Jul 25, 2008 6:53 pm UTC
Location: Montreal

Re: First Tesla Autopilot Death

Postby Chen » Tue Jul 05, 2016 5:36 pm UTC

In terms of maintenance, presumably they will work similarly to aircraft and the time-limited dispatch approach they have to various faults. If there's a severe enough fault with sensors you'd say that the car is not dispatchable (autonomously), until that fault is repaired. Whether that means replacing a sensor or a CPU, you'd have a sort of message in the system and blocks that prevent the car from working in autonomous mode with known faults. If the fault is less severe (single failure of a dual or triple redundant system for example) then maybe you get 50 or 100 hours of operation before you must replace the part. After that time, the car again gets locked out of autonomous mode until the repair is done.

As long as your faults are detected and announced well enough, you're going to end up in similar situations as now. No one is going to willingly drive with failed brakes because of the ridiculous danger it is to THEMSELVES. If your visual sensors are failed, you'd be insane to want to run in autonomous mode willingly anyways.

User avatar
LaserGuy
Posts: 4516
Joined: Thu Jan 15, 2009 5:33 pm UTC

Re: First Tesla Autopilot Death

Postby LaserGuy » Tue Jul 05, 2016 6:16 pm UTC

cphite wrote:
HES wrote:The most serious collisions are when a HGV veers sideways across three lanes, through the central barrier, and across another three lanes of oncoming traffic. Most commonly attributed to tired, distracted, or unwell drivers. Taking that risk away sounds like a great idea to me.


Sure; if you're actually taking that risk away. Not everybody is convinced that it's being taken away.

We already have an example here of an autonomous vehicle failing to notice a tractor trailer.


And? I posted just upthread an example of a bus driver failing to notice a tractor trailer. I think you underestimate how terrible some human drivers are.

Pedestrians and other people (and drivers) is one of the hard parts. Another hard part is that we aren't anywhere close to having perfect sensors that work all the time, or perfect computers that work all the time. For things like trains and planes that undergo constant maintenance and monitoring of systems, you can mitigate that... I'm not sure your average private citizen is going to do that.


No, we're never going to have perfect computers or perfect sensors. Will we have computers and sensors that can perform this task significantly better than a human? Almost certainly. Will we have computers and sensors that can perform this task better than my 90 year old grandfather? I would venture that we already do. It's baffling to me that people are demanding perfection from a system that is 1) less than five years old and 2) when the failure rate for humans at the task is pretty high to begin with.

commodorejohn
Posts: 1083
Joined: Thu Dec 10, 2009 6:21 pm UTC
Location: Placerville, CA
Contact:

Re: First Tesla Autopilot Death

Postby commodorejohn » Tue Jul 05, 2016 7:06 pm UTC

LaserGuy wrote:It's baffling to me that people are demanding perfection from a system that is 1) less than five years old and 2) when the failure rate for humans at the task is pretty high to begin with.

That's easily explained. Not many people who grew up looking at BSODs, GSODs, bombs, Abort/Retry/Fails, or Guru Meditations are ever going to really trust a computer.
"'Legacy code' often differs from its suggested alternative by actually working and scaling."
- Bjarne Stroustrup
www.commodorejohn.com - in case you were wondering, which you probably weren't.

Tyndmyr
Posts: 11028
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: First Tesla Autopilot Death

Postby Tyndmyr » Tue Jul 05, 2016 7:17 pm UTC

I grew up with all of those, and developed a very healthy trust for computers. Yknow, within their limitations, and configured correctly, they are quite reliable. I've run servers that quietly, reliably do their job day in and day out with almost no maint.

Finding people like that is difficult in the extreme.

KnightExemplar
Posts: 5494
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: First Tesla Autopilot Death

Postby KnightExemplar » Tue Jul 05, 2016 8:02 pm UTC

LaserGuy wrote:No, we're never going to have perfect computers or perfect sensors. Will we have computers and sensors that can perform this task significantly better than a human? Almost certainly.


As noted in one of my earlier posts: Tesla has one-death in 130 Million miles driven. In comparison, drivers in general have roughly one-death every 100 Million miles driven.

Including drunk drivers, improper seatbelts, and freak accidents like airbags turning into shrapnel and killing someone. Not all of those accidents were human error to reach 1-death every 100 million miles.

And this "Tesla Autopilot" is an optional, beta system that is only being used for highway driving, and is supposed to be used with an attentive driver ready to take over "at any time" (because its still buggy as shit)

No. This system is not even as good as the crappy humans we are. Maybe one day in the future we will have autonomous drivers that are better than us. But lets not drink the kool-aid and pretend that the day has come yet.
First Strike +1/+1 and Indestructible.

Tyndmyr
Posts: 11028
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: First Tesla Autopilot Death

Postby Tyndmyr » Tue Jul 05, 2016 8:20 pm UTC

KnightExemplar wrote:
LaserGuy wrote:No, we're never going to have perfect computers or perfect sensors. Will we have computers and sensors that can perform this task significantly better than a human? Almost certainly.


As noted in one of my earlier posts: Tesla has one-death in 130 Million miles driven. In comparison, drivers in general have roughly one-death every 100 Million miles driven.

Including drunk drivers, improper seatbelts, and freak accidents like airbags turning into shrapnel and killing someone. Not all of those accidents were human error to reach 1-death every 100 million miles.


The overall system of car+driver kills more than car+robot+person.

Given the caveats of small ass sample size for car+robot+person, of course. The fact that humans get drunk, and then opt to drive...those are legitimate failings of humans. The robot won't do that. That's not something we should be holding against self driving cars.

LaserGuy is entirely correct in that we'll almost certainly have computers and sensors that can perform this task significantly better than a human. Right now, we have a little evidence against this system being that...yet. But we are deriving data from a *single* accident. There's large potential for error there. And obviously there will be improvements.

No. This system is not even as good as the crappy humans we are. Maybe one day in the future we will have autonomous drivers that are better than us. But lets not drink the kool-aid and pretend that the day has come yet.


It's doing pretty good. It's better than some drivers, certainly. All, or even a majority? Maybe not, but it's at least playing in the same rough ballpark. I know everyone believes themselves to be an exceptional driver, and thus, views reliability approaching that of an average driver as crap compared to their mad skillz, but most of those people are wrong. Sure, sure, maybe we COULD be much better, but much of the time we are tired, or drunk, or looking at cell phones, or bored and inattentive. This accident is not so different from what people do.

commodorejohn
Posts: 1083
Joined: Thu Dec 10, 2009 6:21 pm UTC
Location: Placerville, CA
Contact:

Re: First Tesla Autopilot Death

Postby commodorejohn » Tue Jul 05, 2016 8:25 pm UTC

I'm curious about that "130 million miles driven" - that is 130 million miles with the autopilot engaged, right?
"'Legacy code' often differs from its suggested alternative by actually working and scaling."
- Bjarne Stroustrup
www.commodorejohn.com - in case you were wondering, which you probably weren't.

KnightExemplar
Posts: 5494
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: First Tesla Autopilot Death

Postby KnightExemplar » Tue Jul 05, 2016 8:28 pm UTC

commodorejohn wrote:I'm curious about that "130 million miles driven" - that is 130 million miles with the autopilot engaged, right?


Correct.

https://www.teslamotors.com/blog/tragic-loss

This is the first known fatality in just over 130 million miles where Autopilot was activated.
Last edited by KnightExemplar on Tue Jul 05, 2016 8:29 pm UTC, edited 2 times in total.
First Strike +1/+1 and Indestructible.


Return to “News & Articles”

Who is online

Users browsing this forum: No registered users and 20 guests