Aesma wrote:

Why would a remote driver be more ready to take over ?

They're thinking forward to the time when the car has no steering wheel etc.

I think something that could help tremendously is data collection on a large scale of vehicles' positioning. The EU made mandatory for all new cars to have an SOS button that will call assistance and provide GPS location, so all cars will now be GPS enabled and able to provide their location. If that data, let's say in an anonymized form, was transmitted in real time, then autonomous cars tapping into that info would know where other cars have driven previously and that would help with many situations.

Interesting. My car has no built-in GPS, nor any network interface that would allow it to upload such data. I like it that way. Once my car has a network interface, I presume it will be hacked, which is a bad thing.

In case you're interested in the opinion of Top Gear's James May:

To be absolutely clear: something like a truly autonomous car will be with us one day, and that’s great. Developing them is a moral imperative. But we should stop imagining they’re just a few years off.

Ref: https://twitter.com/MrJamesMay/status/9 ... 0204282880

I found https://www.theverge.com/2017/10/10/164 ... i-robotaxi to be pretty interesting.

First, a picture of all the stuff that is needed to process all the data:



Fusion was Uber's previous generation car so we can imagine the new stuff takes up less room, but the article also says that the vendors are now using larger cars (SUVs, wagons) so they have the space for all the network gear, computing, storage, power conditioning and cooling needed to make it all work.

The article also says:

The platform then uses an AI algorithm-based operating system and a cloud-based, high-definition 3D map to help the car understand its environment, know its location, and anticipate potential hazards while driving. The system’s software can be updated over the air — similar to how a smartphone’s operating system is updated — making the car become smarter over time.

Two areas the hackers will be focusing on, IMHO.

Also https://arstechnica.com/cars/2018/01/dr ... explained/ is a very good, readable article about lidar's past, present and future.

A colleague of mine is working on a related application with Cisco, BMW and Valeo : autonomous valet service. It relies on cameras and an nvidia graphic card, but also wifi, because most of that hardware is not in the car but on the premises, in this case my workplace (a large campus of about 1Km² with multiple buildings and parkings).

Aesma wrote:

A colleague of mine is working on a related application with Cisco, BMW and Valeo : autonomous valet service. It relies on cameras and an nvidia graphic card, but also wifi, because most of that hardware is not in the car but on the premises, in this case my workplace (a large campus of about 1Km² with multiple buildings and parkings).

Interesting. It suggests that maybe fully autonomous driving is a bit further out than planned.

NY Times review of Tesla Model 3 ( https://www.nytimes.com/2018/04/05/busi ... del-3.html ) sums the situation up as:

At this juncture, Autopilot is a misnomer, as many others have observed. “This is not self-driving technology,” said Bryan Reimer, a research scientist and expert on autonomous vehicle technology at the Massachusetts Institute of Technology. “It’s a collaborative system that leaves responsibility for safety in the drivers’ hands and isn’t intended to let them sit in the back seat or pursue other activities while driving.” Even though some companies have said they plan driverless ride services in as little as two years, Mr. Reimer said full automation was likely to be many years or even decades in the future.

On the positive side:

After my drive in the Model 3, my sense is that even with its flaws, Tesla’s Autopilot is a vast improvement over conventional controls, and enhances (rather than supplants) the driving experience. I’m tempted to join the Model 3 waiting list, which numbers over 400,000 customers.

Well it appears the system in the Uber car did in fact "see" Ms. Herzberg before it struck her:

The cause of the fatal crash of an Uber self-driving car appears to have been at the software level, specifically a function that determines which objects to ignore and which to attend to, The Information reported.
[...]

The sources cited by The Information say that Uber has determined B was the problem. Specifically, it was that the system was set up to ignore objects that it should have attended to; Herzberg seems to have been detected but considered a false positive.

https://techcrunch.com/2018/05/07/uber- ... it-struck/

Nothing else is being said so far as this is apparently inside information but importantly no one is denying it (plus it makes sense, it follows what has been postulated)

Tugg

Tugger wrote:

Well it appears the system in the Uber car did in fact "see" Ms. Herzberg before it struck her:

The cause of the fatal crash of an Uber self-driving car appears to have been at the software level, specifically a function that determines which objects to ignore and which to attend to, The Information reported.
[...]

The sources cited by The Information say that Uber has determined B was the problem. Specifically, it was that the system was set up to ignore objects that it should have attended to; Herzberg seems to have been detected but considered a false positive.

https://techcrunch.com/2018/05/07/uber- ... it-struck/

Nothing else is being said so far as this is apparently inside information but importantly no one is denying it (plus it makes sense, it follows what has been postulated)

Tugg

Apparently the car noticed her 6 seconds before impact, but wasn't sure what to classify her as, which led to confusion on the part of the software. Then things got worse: the car determined a need for emergency braking 1.3 seconds before impact, but that feature had been disabled in order to prevent accidental activation which would have adversely affected passenger comfort. It was instead up to the driver to take control, however the car was not designed to alert the driver to the need for this. The car itself came with its own emergency braking function, but Uber disabled it.

In addition, Uber required drivers to monitor a screen in the car in order to keep an eye on diagnostic messages to flag those of interest for further review. Uber used to have two people in the car so that one could monitor the road while the other monitored the screen, but to save costs the two jobs were combined to be done by one person.

https://arstechnica.com/cars/2018/05/em ... ntsb-says/

Whoever designed this thing should find a new line of work that doesn't involve safety-critical systems. Whoever allowed it to go out on the road in a fail-deadly state like that should face some negligence charges. And Uber should be looking at a huge liability suit.

A Tesla sedan in Autopilot mode crashed into a parked Laguna Beach Police Department vehicle Tuesday morning, authorities said.
The collision happened at 11:07 a.m. at 20652 Laguna Canyon Road, according to Laguna Police Sgt. Jim Cota. The officer was not in the cruiser at the time of the crash. The Tesla driver suffered minor injuries, but refused transportation to the hospital.
"Thankfully there was not an officer at the time in the police car," Cota said. "The police car is totaled."
Cota said that a year ago in the same area there was another collision involving a Tesla running into a semi-truck.

http://www.latimes.com/local/lanow/la-m ... story.html

Another case where the driver wasn't following the rules while the vehicle was in the autopilot mode.

910A wrote:

A Tesla sedan in Autopilot mode crashed into a parked Laguna Beach Police Department vehicle Tuesday morning, authorities said.
The collision happened at 11:07 a.m. at 20652 Laguna Canyon Road, according to Laguna Police Sgt. Jim Cota. The officer was not in the cruiser at the time of the crash. The Tesla driver suffered minor injuries, but refused transportation to the hospital.
"Thankfully there was not an officer at the time in the police car," Cota said. "The police car is totaled."
Cota said that a year ago in the same area there was another collision involving a Tesla running into a semi-truck.

http://www.latimes.com/local/lanow/la-m ... story.html

Another case where the driver wasn't following the rules while the vehicle was in the autopilot mode.

https://arstechnica.com/cars/2018/05/au ... skyrocket/ tells us the Tesla Model S is now the most expensive car to insure.

Not a big surprise to me at least.

The insurance industry is a lot harder to bamboozle than the media is.

Revelation wrote:

At this juncture, Autopilot is a misnomer, as many others have observed. “This is not self-driving technology,” said Bryan Reimer, a research scientist and expert on autonomous vehicle technology at the Massachusetts Institute of Technology. “It’s a collaborative system that leaves responsibility for safety in the drivers’ hands and isn’t intended to let them sit in the back seat or pursue other activities while driving.” Even though some companies have said they plan driverless ride services in as little as two years, Mr. Reimer said full automation was likely to be many years or even decades in the future.

Dumb people are allowed on the roads. Senile people are allowed on the roads. Drunk people are allowed on the roads. There is no way the tech is decades away from prime time. 5 years at most.

Aesma wrote:

Dumb people are allowed on the roads. Senile people are allowed on the roads. Drunk people are allowed on the roads. There is no way the tech is decades away from prime time. 5 years at most.

If you're accepting Tesla statements at face value, you're the dumb/senile/drunk person here.

As the old saying goes, never ask a barber if he thinks you need a haircut.

Their customer base is not a cross section of the general population, yet they keep acting as if it is. It's the oldest trick in the book.

These vehicles aren't going to replace the average dummy driver, they're going to go to those who can afford them, which in general means people who already have their act together and in general are driving in well developed areas and in general will maintain the vehicle to high standards.

Even still this idealized clientele are showing the flaws in the technology.

It's not just Tesla working on this. Many companies are. Here the thread is about Uber, for example.

JJJ wrote:

Aesma wrote:

It's not just Tesla working on this. Many companies are. Here the thread is about Uber, for example.

Tesla is actually bottom of the pack.

Point is you should try to find someone who isn't part of a company making or selling the tech to get a less biased read on what is going on.

For instance, up thread I quoted James May as saying the tech is coming but just not as soon as many think it is, and I think that's a good read on the situation.

I think car should read out disclaimer every time driver turns on autonomous mode.

Also when tech cannot handle a particular situation it should say "Your Car Mr/Mrs/Ms ...".

frmrCapCadet wrote:

Let me restate what happened. Other posters have said it, but here is what pops:

The Volvo had a pedestrian sensing system and automatic braking. Uber disabled it. The system did detect the pedestrian and would have worked.

That is as believable as amazon's explanation on why Alexa e-mailed recorded private conversation to others.

While in autonomous mode every braking is automatic braking, including when you reach your intended destination. Otherwise what is the plan, jump out of the window because automatic braking is disabled.

As if we didn't know already, https://tech.slashdot.org/story/18/06/2 ... ays-police tells us the flunky that Uber was paying to monitor the self-driving car was watching Hulu instead.

Police obtained records from Hulu suggesting that Vasquez was watching "The Voice," a singing talent competition that airs on NBC, just before the crash. Hulu's records showed she began watching the program at 9:16pm. Streaming of the show ended at 9:59pm, which "coincides with the approximate time of the collision," according to the police report.

Lesson to be learned: humans are unreliable, and any scheme that relies on them to pay attention to a monotonous stream of information is bound to fail.

Full article ( https://arstechnica.com/cars/2018/06/po ... car-crash/ ) says:

While the report is bad news for Vasquez, the NTSB report also found that poor engineering decisions by Uber contributed to the crash. The NTSB report found that Uber's software "determined that an emergency braking maneuver was needed" 1.3 seconds before the crash. Unfortunately, the vehicle wasn't programmed to actually perform emergency braking procedures—nor was it programmed to alert the safety driver.

So no one involved covered themselves in glory here.

Interesting contribution for the argument why two pilots are required in the cockpit....

Tugg

Revelation wrote:

As if we didn't know already, https://tech.slashdot.org/story/18/06/2 ... ays-police tells us the flunky that Uber was paying to monitor the self-driving car was watching Hulu instead.

Police obtained records from Hulu suggesting that Vasquez was watching "The Voice," a singing talent competition that airs on NBC, just before the crash. Hulu's records showed she began watching the program at 9:16pm. Streaming of the show ended at 9:59pm, which "coincides with the approximate time of the collision," according to the police report.

Lesson to be learned: humans are unreliable, and any scheme that relies on them to pay attention to a monotonous stream of information is bound to fail.

Full article ( https://arstechnica.com/cars/2018/06/po ... car-crash/ ) says:

While the report is bad news for Vasquez, the NTSB report also found that poor engineering decisions by Uber contributed to the crash. The NTSB report found that Uber's software "determined that an emergency braking maneuver was needed" 1.3 seconds before the crash. Unfortunately, the vehicle wasn't programmed to actually perform emergency braking procedures—nor was it programmed to alert the safety driver.

So no one involved covered themselves in glory here.

Thanks for the Link Rev, I caught something on the news the other day about it but forgot about it. On the in car camera he is defiantly not looking out the window and down at something, I am not sure due to the poor video he would have seen her under the street lamps.

As I have posted above I have the joy of working on this stuff at my dealership, this week alone have done 2 Re-calibrations of the forward collision warning Multipurpose camera, because of windshields replaced. It works in conjunction under some conditions with the Radar behind the Bumper.
I did also hopefully repair a customers car.... complaining that randomly it would brake hard with no car stopping in Front of her, with a bit more prying with some questions it seems to happen when there was a dip in the road but not every time... with a bit more looking at the car and starting to re-calibrate the Radar behind the Left Front. bumper I saw some damage to the under tray, seems she climbed a curb stone and it pushed up and back on the lower bracket that supports the radar causing it to point down just s small amount, I am thinking enough to cause it reflect off the road surface coming up as the cars front end dropped into the dip. I replaced the Bracket and The Radar, re calibrated and no complaints have been reported yet.....

C

Tugger wrote:

Still a lot more investigation ahead and this will shape the policies for future autonomous regulation. One thing I noted is that the speed limit was 35 and the car was going 38, while that may be "normal" for a lot of human drivers I think autonomous cars will be required to hew to the speed limit.

Tugg

You completely over looked some very basic facts. The reaction time for a human is something like 3/4 to 1 1/2 seconds were as the reaction time for an autonomous car is measured in Milliseconds and also an autonomous car does not get distracted or impaired. If anything, speed limit could be raised considerably and following distance could be reduced considerably.

The unfortunate ladies chances of survival may have actually increased substantially because the car was not being driven By a human. The idea of a "Pilot Driver" is a pretty ridiculous idea because it is found that the Pilot Drivers reaction time becomes longer than when they are actually driving the car. The biggest problem I see with bike riders is that they think that they are privileged and the rules of the road (or common sense) don't apply to them.

There is a lot of work still to do in writing the software and control laws for road transportation but we will get their. Airbus had a shaky start with automation in the beginning and has greatly improved safety and implementation since then. https://www.youtube.com/watch?v=KEH7OpnA-I4

From Wikipedia https://en.wikipedia.org/wiki/Autonomous_car
Traffic
Additional advantages could include higher speed limits;[94] smoother rides;[95] and increased roadway capacity; and minimized traffic congestion, due to decreased need for safety gaps and higher speeds.[96][97] Currently, maximum controlled-access highway throughput or capacity according to the U.S. Highway Capacity Manual is about 2,200 passenger vehicles per hour per lane, with about 5% of the available road space is taken up by cars. One study estimated that autonomous cars could increase capacity by 273% (~8,200 cars per hour per lane). The study also estimated that with 100% connected vehicles using vehicle-to-vehicle communication, capacity could reach 12,000 passenger vehicles per hour (up 445% from 2,200 pc/h per lane) traveling safely at 120 km/h (75 mph) with a following gap of about 6 m (20 ft) of each other. Currently, at highway speeds drivers keep between 40 to 50 m (130 to 160 ft) away from the car in front. These increases in highway capacity could have a significant impact in traffic congestion, particularly in urban areas, and even effectively end highway congestion in some places.[98] The ability for authorities to manage traffic flow would increase, given the extra data and driving behavior predictability.[7] combined with less need for traffic police and even road signage.

stratclub wrote:

If anything, speed limit could be raised considerably and following distance could be reduced considerably.

Sure, once the hardware and software are up to snuff, and EVERY non-autonomous car is eliminated.

Till then we're stuck with the least common denominator.

Revelation wrote:

stratclub wrote:

If anything, speed limit could be raised considerably and following distance could be reduced considerably.

Sure, once the hardware and software are up to snuff, and EVERY non-autonomous car is eliminated.

Till then we're stuck with the least common denominator.

I agree that the tech needs some work.

I don't think all other cars need to be taken off of the road first. How about turning carpool lanes into self driving car lanes only complete with their own speed limit? Could work being that they are doing trials right now for variable speed limit lanes in Seattle.
http://www.wsdot.wa.gov/smarterhighways/vsl.htm

stratclub wrote:

Revelation wrote:

stratclub wrote:

If anything, speed limit could be raised considerably and following distance could be reduced considerably.

Sure, once the hardware and software are up to snuff, and EVERY non-autonomous car is eliminated.

Till then we're stuck with the least common denominator.

I agree that the tech needs some work.

I don't think all other cars need to be taken off of the road first. How about turning carpool lanes into self driving car lanes only complete with their own speed limit? Could work being that they are doing trials right now for variable speed limit lanes in Seattle.
http://www.wsdot.wa.gov/smarterhighways/vsl.htm

Interesting idea, but as you see it's to address a different class of problem.

The problem with special lanes for autonomous vs non-autonomous is how do vehicles enter/exit the flow without disrupting it to the point of reducing its benefits (e.g. +275% capacity gain given above). How do you merge traffic into and out of the flow without slowing it down or providing for very long ingress/egress lanes?

I know in the Hartford CT area they have special HOV lanes from the Eastern suburbs on I-84 but they all terminate in the city. Basically you'd prefer the autonomous pack setup to be an express lane bypassing the city, because the main problem is that the city can't absorb the current inflow never mind an inflow increased by ~+275%.

stratclub wrote:

Tugger wrote:

Still a lot more investigation ahead and this will shape the policies for future autonomous regulation. One thing I noted is that the speed limit was 35 and the car was going 38, while that may be "normal" for a lot of human drivers I think autonomous cars will be required to hew to the speed limit.

Tugg

You completely over looked some very basic facts. The reaction time for a human is something like 3/4 to 1 1/2 seconds were as the reaction time for an autonomous car is measured in Milliseconds and also an autonomous car does not get distracted or impaired. If anything, speed limit could be raised considerably and following distance could be reduced considerably.

The unfortunate ladies chances of survival may have actually increased substantially because the car was not being driven By a human. The idea of a "Pilot Driver" is a pretty ridiculous idea because it is found that the Pilot Drivers reaction time becomes longer than when they are actually driving the car. The biggest problem I see with bike riders is that they think that they are privileged and the rules of the road (or common sense) don't apply to them.

There is a lot of work still to do in writing the software and control laws for road transportation but we will get their. Airbus had a shaky start with automation in the beginning and has greatly improved safety and implementation since then. https://www.youtube.com/watch?v=KEH7OpnA-I4

From Wikipedia https://en.wikipedia.org/wiki/Autonomous_car
Traffic
Additional advantages could include higher speed limits;[94] smoother rides;[95] and increased roadway capacity; and minimized traffic congestion, due to decreased need for safety gaps and higher speeds.[96][97] Currently, maximum controlled-access highway throughput or capacity according to the U.S. Highway Capacity Manual is about 2,200 passenger vehicles per hour per lane, with about 5% of the available road space is taken up by cars. One study estimated that autonomous cars could increase capacity by 273% (~8,200 cars per hour per lane). The study also estimated that with 100% connected vehicles using vehicle-to-vehicle communication, capacity could reach 12,000 passenger vehicles per hour (up 445% from 2,200 pc/h per lane) traveling safely at 120 km/h (75 mph) with a following gap of about 6 m (20 ft) of each other. Currently, at highway speeds drivers keep between 40 to 50 m (130 to 160 ft) away from the car in front. These increases in highway capacity could have a significant impact in traffic congestion, particularly in urban areas, and even effectively end highway congestion in some places.[98] The ability for authorities to manage traffic flow would increase, given the extra data and driving behavior predictability.[7] combined with less need for traffic police and even road signage.

The very basic fact is, the car should have seen her ( it did) and stopped.( It did not)
Why? Now we know... because it wasn't programmed to stop according to NTSB... and now you want to speed things up, that's nuts.
More than a lot of Software work to be done here, I know this is whats coming but there is such a mix of cars that are almost self driving and people driven that I think it will get worse before it gets better.

Aesma wrote:

A colleague of mine is working on a related application with Cisco, BMW and Valeo : autonomous valet service. It relies on cameras and an nvidia graphic card, but also wifi, because most of that hardware is not in the car but on the premises, in this case my workplace (a large campus of about 1Km² with multiple buildings and parkings).

graphics card for computing oomphf? ( ref: https://en.wikipedia.org/wiki/CUDA )

I do wonder when the first major pileup happens because someone hacked the cloud based geo information with a non existent tunnel through a building or such.

Classa64 wrote:

stratclub wrote:

Tugger wrote:

Still a lot more investigation ahead and this will shape the policies for future autonomous regulation. One thing I noted is that the speed limit was 35 and the car was going 38, while that may be "normal" for a lot of human drivers I think autonomous cars will be required to hew to the speed limit.

Tugg

You completely over looked some very basic facts. The reaction time for a human is something like 3/4 to 1 1/2 seconds were as the reaction time for an autonomous car is measured in Milliseconds and also an autonomous car does not get distracted or impaired. If anything, speed limit could be raised considerably and following distance could be reduced considerably.

The unfortunate ladies chances of survival may have actually increased substantially because the car was not being driven By a human. The idea of a "Pilot Driver" is a pretty ridiculous idea because it is found that the Pilot Drivers reaction time becomes longer than when they are actually driving the car. The biggest problem I see with bike riders is that they think that they are privileged and the rules of the road (or common sense) don't apply to them.

There is a lot of work still to do in writing the software and control laws for road transportation but we will get their. Airbus had a shaky start with automation in the beginning and has greatly improved safety and implementation since then. https://www.youtube.com/watch?v=KEH7OpnA-I4

From Wikipedia https://en.wikipedia.org/wiki/Autonomous_car
Traffic
Additional advantages could include higher speed limits;[94] smoother rides;[95] and increased roadway capacity; and minimized traffic congestion, due to decreased need for safety gaps and higher speeds.[96][97] Currently, maximum controlled-access highway throughput or capacity according to the U.S. Highway Capacity Manual is about 2,200 passenger vehicles per hour per lane, with about 5% of the available road space is taken up by cars. One study estimated that autonomous cars could increase capacity by 273% (~8,200 cars per hour per lane). The study also estimated that with 100% connected vehicles using vehicle-to-vehicle communication, capacity could reach 12,000 passenger vehicles per hour (up 445% from 2,200 pc/h per lane) traveling safely at 120 km/h (75 mph) with a following gap of about 6 m (20 ft) of each other. Currently, at highway speeds drivers keep between 40 to 50 m (130 to 160 ft) away from the car in front. These increases in highway capacity could have a significant impact in traffic congestion, particularly in urban areas, and even effectively end highway congestion in some places.[98] The ability for authorities to manage traffic flow would increase, given the extra data and driving behavior predictability.[7] combined with less need for traffic police and even road signage.

The very basic fact is, the car should have seen her ( it did) and stopped.( It did not)
Why? Now we know... because it wasn't programmed to stop according to NTSB... and now you want to speed things up, that's nuts.
More than a lot of Software work to be done here, I know this is whats coming but there is such a mix of cars that are almost self driving and people driven that I think it will get worse before it gets better.

The very basic fact is that she did something that defies common sense, she J-walked into lanes of traffic from between some trees and stepped out into traffic without looking. If a human was driving the car, would the outcome have been any different? I would say that Darwin did the right thing and took her out of the gene pool. What was the amount of time between when the person was spotted and the collision occurred? Milliseconds maybe?

stratclub wrote:

Classa64 wrote:

stratclub wrote:

You completely over looked some very basic facts. The reaction time for a human is something like 3/4 to 1 1/2 seconds were as the reaction time for an autonomous car is measured in Milliseconds and also an autonomous car does not get distracted or impaired. If anything, speed limit could be raised considerably and following distance could be reduced considerably.

The unfortunate ladies chances of survival may have actually increased substantially because the car was not being driven By a human. The idea of a "Pilot Driver" is a pretty ridiculous idea because it is found that the Pilot Drivers reaction time becomes longer than when they are actually driving the car. The biggest problem I see with bike riders is that they think that they are privileged and the rules of the road (or common sense) don't apply to them.

There is a lot of work still to do in writing the software and control laws for road transportation but we will get their. Airbus had a shaky start with automation in the beginning and has greatly improved safety and implementation since then. https://www.youtube.com/watch?v=KEH7OpnA-I4

From Wikipedia https://en.wikipedia.org/wiki/Autonomous_car
Traffic
Additional advantages could include higher speed limits;[94] smoother rides;[95] and increased roadway capacity; and minimized traffic congestion, due to decreased need for safety gaps and higher speeds.[96][97] Currently, maximum controlled-access highway throughput or capacity according to the U.S. Highway Capacity Manual is about 2,200 passenger vehicles per hour per lane, with about 5% of the available road space is taken up by cars. One study estimated that autonomous cars could increase capacity by 273% (~8,200 cars per hour per lane). The study also estimated that with 100% connected vehicles using vehicle-to-vehicle communication, capacity could reach 12,000 passenger vehicles per hour (up 445% from 2,200 pc/h per lane) traveling safely at 120 km/h (75 mph) with a following gap of about 6 m (20 ft) of each other. Currently, at highway speeds drivers keep between 40 to 50 m (130 to 160 ft) away from the car in front. These increases in highway capacity could have a significant impact in traffic congestion, particularly in urban areas, and even effectively end highway congestion in some places.[98] The ability for authorities to manage traffic flow would increase, given the extra data and driving behavior predictability.[7] combined with less need for traffic police and even road signage.

The very basic fact is, the car should have seen her ( it did) and stopped.( It did not)
Why? Now we know... because it wasn't programmed to stop according to NTSB... and now you want to speed things up, that's nuts.
More than a lot of Software work to be done here, I know this is whats coming but there is such a mix of cars that are almost self driving and people driven that I think it will get worse before it gets better.

The very basic fact is that she did something that defies common sense, she J-walked into lanes of traffic from between some trees and stepped out into traffic without looking. If a human was driving the car, would the outcome have been any different? I would say that Darwin did the right thing and took her out of the gene pool. What was the amount of time between when the person was spotted and the collision occurred? Milliseconds maybe?

Defies common sense? So you have never crossed the street outside a controlled intersection or cross walk , come on....
You 100% right she did cross the street by J walking and a million people do it everyday, some get hit even when looking and some don't. My point was there all about how safe these cars are and how much better they are than us, how much faster.
I agree in ways they are, they will save lives it also looks like they take them to, in the name of progress I think there is more work to be done. At this time in self driving cars its going to be all about who is going to sue who... so lady J walks illegally , maybe she did look, maybe she was drunk, hell maybe she wanted to die and she gets hit by a self driving car that should have avoided her but chose not to... who's at fault.