I think that’s what the conclusion is coming to right now. Unless Congress or state governments toughen road laws and make them stick across the board I think adding automated cars into the mix of traffic is a bad idea even if it’s safer by themselves (unless we ban manual driver altogether which is another solution which might happen).
3 Likes
Well heck, it should only take another what? 5 to 10 pedestrian deaths for their actuaries to notice?
Here is another perspective on the tech from someone who has some experience with autopilot, and having hundreds of lives in his hands on the regular.
https://ftalphaville.ft.com/2017/03/15/2186054/snap-av-captain-sully-on-autonomous-cars/
2 Likes
So what your image shows is that a pedestrian in a black shirt with no reflectors decided to cross a 5-lane road rather than walk 100 feet and cross at the cross walk. Also, if you were crossing a 1-way street, why would you not place the bike between yourself and oncoming traffic “just in case”.
Just recently I have driven down Central in Phoenix and narrowly missed a bicyclist. They were riding their bike on the cross walk against the light. I have good night vision and am young enough to still have good reaction time. A driver with poor night vision or slower reaction time would have hit them.
Neither the autonomous Uber or the pedestrian are 100% at fault. I think it is about a 60/40 split.
1 Like
If the end result is that autonomous cars become safe and common sooner than they otherwise would, therefore saving more than 5-10 pedestrian lives, then I’d be very happy with that outcome. I’d be happier to avoid those deaths as well, if possible.
There’s a lot of room on the spectrum between unregulated libertarian dystopia and overregulated police state/nanny state, room which includes the FDA, the TSA, the DMV, and so on. We seem to prefer slightly different tradeoffs in risk tolerance. I lean towards “take whichever paths lead to the least total human suffering regardless of cause.” Many others seem to lean towards “take whichever paths lead to the least human suffering attributable to novel causes.”
does that split which is totally yours to determine make this less scary?
i really cant imagine any other reason to say such a thing.
remember folks, some people on the internet are invested in this. there is a lot of money at stake, and thats before we consider the egos at play.
sp3, you and i disagree on the value of a human life and what our responsibilities to one another are.
1 Like
I agree, that looks really bright. It may have been the same location. Was it the same moon phase? Same cloud coverage?
You know, riding your bike in a crosswalk is illegal…
*Citation needed
I thought the whole point of robot drivers was that they would be better and safer than slow, distractible, falliable human drivers. Unless the entire auronomous vehicle craze sweeping thr tech indistry is just a bunch of overhyped hot air.
Obviously that’s the promise, however, expecting this technology to achieve perfection in its infancy is pretty damn naive. In this situation the vehicle appears to have performed about on par with a human driver, that’s already pretty incredible.
So doing better than a human most of the time and only as good as a human some of the time is a “fail”? People in rural areas are in this position frequently when deer cross the road, and deer are frequently hit. Pedestrians cannot simply obliviously walk into traffic confident vehicles will see them and be able to stop. A human driver may have been looking in his rear view or changing the radio station, one can’t assume 100% attentiveness.
Why would you assume that? These are complex systems that seem to be functioning quite well, as said we’d all love to see the miles vs accidents stats compared to humans
This is key, risk acceptance is directly tied to perception of control. There are people happy to drive somewhere rather than fly, even though statistically they are far safer flying. Self driving cars is surrendering control of your car. But as frequent observer of bad drivers, as a pedestrian, cyclist or other driver on the road, the bar is low to be better than them.
Question for tech types: How hard would it be for the car to sense rf transmitting devices like cell phones as a safety feature? Odd are very good most people over 12 have one on their person. I know they keep in touch with the tower, but I don’t know how frequently.
.
1 Like
They weren’t at the T intersection though, they were about 100 feet from it.
2 Likes
Actually, many, if not most states tell bicyclists to dismount, when crossing any road, other than when riding down a cross street, and it’s mandatory (for anyone over 12, usually) when using sidewalks/crosswalks. A pedestrian has MUCH better acceleration than anyone on a bicycle, for one, and you can jump backwards (or to the side).
1 Like
It doesn’t make it less scary. I also fully realize this is my own opinion.
Driving in its self is scary: “lets get some big hunks of metal moving really fast” - then you introduce a plethora of outside variables that the driver can not control (other drivers, pedestrians, wildlife, weather…)
However, I take a look at distracted drivers, constantly on their cellphones and I think - is this really that much worse. This is just being reported and blowing up because its new and “scary” tech that was involved in a death.
1 Like
Did those studies include the costs to individuals who didn’t knowingly agree to be in the “trial” having adverse affects? Since, to them, the personal cost is way off the scale. Or, did they include labeling requirements for putting stuff on the shelf equivalent to the documents actual trial subjects have to agree to and sign along with requiring signature to actually buy them?
It may be statistically better for the whole to run trials by simply releasing products and recalling the bad ones instead of proving safety (mostly) and efficacy (mostly) ahead of time. But, at the individual level the risks and costs aren’t the same.
I prefer knowing that something on the shelf is at least not supposed to harm me. If that was to change, we would need to end up with something similar still done for labeling requirements. That way we know it’s all just a crap shoot and we’re just participants in the experiment.
Otherwise, we’re just back to tonic water with random extra ingredients being sold as a cure all for anything. No way of knowing if it’s effective or safe until after the recalls start happening and random people have paid the cost.
Back to the driverless car. I thought the car systems passed some level of closed course testing prior to being released onto public roads. In this case, while the video shows a driver having a hard time seeing the person, the entire point of the driverless car is that it has better senses than a human driver. If it doesn’t have better senses, better reaction time, better concentration, better everything, then what’s the point. This looks like a total failure of the system. Send it back to the closed course trials until it’s really ready to come back.
4 Likes
1 Like
There’s no uniform cycling code in Arizona. In Tempe it only states that cyclists riding in a cross walk continue to be considered vehicles and give up the right of way granted to pedestrians. Furthermore cyclists may use sidewalks except where prohibited by signage.
1 Like
I made no specific claims for AZ ^^'. I’ll note, however, that if the collision was far enough from the nearest intersection, the bicyclist wasn’t even breaking any potential jaywalking laws (the dark outfit was, ofc, quite silly).
There really is no excuse; LIDAR should not be susceptible to this kind of problem. Furthermore, the bicycle had moving reflectors; those should trigger ANY vision system.
Have you ever listened to or read any ad for a pharmaceutical? (Whether or not prescription drugs should be advertised is questionable, that’s not the issue)
The list of possible side effects is often the longest part, but there is still no doubt that these drugs do a lot for the health of millions, in the USA and around the world.
When you say “not supposed to harm me” are you suggesting that any drugs which may have any harmful side effects for any single user should be taken off market?
Obligatory on topic comment: Pretty much the same rule applies to autonomous cars. If the legal/liability principle is established as “no pedestrian or cyclist can ever be struck by an autonomous vehicle”, then there just won’t be any operational autonomous vehicles. There will remain all the human driven vehicles, which on balance, will certainly be more dangerous.
Danke. I could’ve looked it up but in the middle of work and multitasking has been difficult today.
