[interested_observer]

Pretty sure that Tesla already is doing this very thing. Their cars are connected to the web, and it would not be a big deal to have embedded code that sends and receives pertinent data to improve the design of the software and the systems it operates.

With vehicles connected to the web in this way, there are security issues. There have already been documented instances where vehicles have been taken over.

• The Jeep Hackers Are Back to Prove Car Hacking Can Get Much Worse | WIRED
• A Deep Flaw in Your Car Lets Hackers Shut Down Safety Features | WIRED
• Hackers Remotely Kill a Jeep on the Highway?With Me in It | WIRED

Yes, very clever. Which is why Telsa owners can push a button on the steering wheel, say "Bug report" followed by the fault they're seeing, or improvement they want, and that is automatically sent to HQ and may then be implemented as an "over the air" update later on.

Over the air updates are fine in theory, but very subject to being hacked. It's much better to have any and all updates performed via a hard connection upload. You really do not want your vehicle to get a new software upload while driving.

As long as it's not dependent on a Windows operating system.

Well, actually Blackberry's QNX is on Ford. Microsoft is a player too....

• https://humanizing.tech/the-new-battleground-car-operating-systems-609592e07a31
• Microsoft Connected Vehicle Platform helps automakers transform cars - The Official Microsoft Blog

For sure.

Most operating systems like those used in vehicle ECM units are Unix based.

... which is a problem. Unix / Linux or various *nix derivatives all have various security problems. BSD Unix was completely re-architected, re-designed and re-written from a clean sheet, with the object of producing a secure *nix baseline to date. Overall it has been very successful. Having said that, it's still not a trusted separation kernel.

The problem is that everyone (the vehicle manufacturers) along with the various OS vendors are using what they have and what is available, just like they always have - and what the world is using is a very insecure, not trustworthy Internet. Security has always been the last band-aid applied, as opposed to being baked into the system from the inception.

This does not bode well for the autonomous vehicle industry - and their occupants/passengers.

I might also add that most manufacturers that offer cars and pickups that are web connected are using bi-directional communications with the ECM, BCM, TCM, etc. on their vehicles. Like Mark mentioned with Tesla, the data is used to improve the product, and updates can even be done without even a visit to the dealer service department.

The concept sounds great in theory, however in actual implementation and operation, it becomes a nightmare.... Without a trusted and secure base system - this is a house of cards waiting to collapse.

... just saying.

[PDL]

I'm not advocating tracking in any full-time, Big Brother sense of the word. Instead, I advocate something more like what's already done in the software industry. For example, most modern smartphones, operating systems, and even some applications send "crash reports" back to the maker of the hardware or software. Some also send diagnostic data on how the system is performing.

You mean like the supposed intrusive tracking that Microsoft is doing with Windows 10? With all the push back, now they are going to track = sell that information, to those other guys?

And, yes, each phone & software company only collects their own data on their own systems using their own proprietary formats. But with tens or even hundreds of millions of devices in operation by each large vendor, they probably get a really good picture of where the weak spots are. Sure, standards would be lovely, but they really aren't needed at all because each of the major car makers is so large, they'll have a huge sample size of event data in no time. Standards would only slow the process down.

Really?

The government agencies I mentioned are even less intrusive. The NTSB only steps if an accident occurs (and they have done so in a number of cases of Tesla crashes). The FAA runs an anonymous reporting system by which any pilot can report an incident or near-miss. Neither agency tracks everything, they just collect data on problems and analyze near-misses and crashes with an eye to improving equipment, maintenance, and procedures. My point is to replicate the process of analyzing near-misses and incidents with the intention of fixing the car's software and rolling it out to the fleet.


Two things, Talk to Boeing about FAA intrusion, I did work there supporting Reliability and Maintainability Engineering and the FAA was plenty intrusive. Second, a near-miss means that two or more objects have actually collided. Sort of like hitting the edge of the target is a near miss. The proper term is near collision not near-miss.

None of that requires 24x7 tracking, only an end-day or end-of-week upload of anomalies and diagnostics .

How is that going to help when you encounter a new situation, or one that that requires communication to the hive? And it can't be using wireless since you have stated that wireless is not a requirement.

[LensBeginner]

Edmonton, Alberta.

I remember a story in an Edmonton paper a few years ago that Edmonton International Airport had to close down one morning because it was -44°C and too dangerous for airport operations. Luckily it had warmed up to -40°C by lunch time and the airport could be reopened. I remember thinking at the time that anyone who can hold warming up and -44°C in the same thought has a very skewed idea of temperature.

...wow...

[Racer X 69]

With vehicles connected to the web in this way, there are security issues. There have already been documented instances where vehicles have been taken over.

• The Jeep Hackers Are Back to Prove Car Hacking Can Get Much Worse | WIRED
• A Deep Flaw in Your Car Lets Hackers Shut Down Safety Features | WIRED
• Hackers Remotely Kill a Jeep on the Highway?With Me in It | WIRED

Over the air updates are fine in theory, but very subject to being hacked. It's much better to have any and all updates performed via a hard connection upload. You really do not want your vehicle to get a new software upload while driving.


Well, actually Blackberry's QNX is on Ford. Microsoft is a player too....

• https://humanizing.tech/the-new-battleground-car-operating-systems-609592e07a31
• Microsoft Connected Vehicle Platform helps automakers transform cars - The Official Microsoft Blog

... which is a problem. Unix / Linux or various *nix derivatives all have various security problems. BSD Unix was completely re-architected, re-designed and re-written from a clean sheet, with the object of producing a secure *nix baseline to date. Overall it has been very successful. Having said that, it's still not a trusted separation kernel.

The problem is that everyone (the vehicle manufacturers) along with the various OS vendors are using what they have and what is available, just like they always have - and what the world is using is a very insecure, not trustworthy Internet. Security has always been the last band-aid applied, as opposed to being baked into the system from the inception.

This does not bode well for the autonomous vehicle industry - and their occupants/passengers.


The concept sounds great in theory, however in actual implementation and operation, it becomes a nightmare.... Without a trusted and secure base system - this is a house of cards waiting to collapse.

... just saying.

I fully agree on every point.

Fortunately for me none of the newer vehicles we own have systems that can be accessed wirelessly.

[Racer X 69]

Where do we find snow or ice at -40°C on earth? the poles? Siberia? so I'm quite optimistic at least regarding the temperature issue

Winnipeg, Manitoba.

I was there a week before Christmas some years back, and the temperatures were -50 F.

The locals told me it is like that every winter.

[MarkJerling]

In 1886, Karl Benz invented the first internal combustion engine driven motor vehicle. 10 years later, Henry Ford built his first car. In 1908 Henry gave the world the Model T and the rest, as they say, is history. No matter how much angst we have about this new direction and autonomous vehicles; it's coming and these cars will be safer and more reliable that any thing we have now. Will there be teething problems? Of course.

I would not get too hung up on hacking. 10 years ago, governments were hacking each other's industrial processes. We have not, as result, shut down our computer networks. It's not going to be an impediment to development. A minor irritation perhaps. But hackers and security experts will continue to square off for years to come.

[Racer X 69]

If accumulated snow would block or otherwise affect sensors this should prove to be very "interesting".
Regularly I see people driving with four inches of snow on the roof and partially cleared (sometimes barely so) windshields.

Chris

Vehicles covered with snow?

Check this out.

Youtu.be

Youtu.be

[Rondec]

I know there is a lot of fear out there about government control and hackers and things like that. Also fear of how these sort of systems will react to unforeseen conditions.

The one thing I would say is that my biggest fear when I get behind the wheel is from other drivers. People drive overly aggressively, or completely distracted and cause significant higher numbers of accidents and traffic fatalities than do bad conditions. And situations like road construction where the road narrows by a lane or two could be handled significantly better in situations where vehicles were able to maintain speed and still merge together (which definitely seldom happens with humans at the helm).

[Racer X 69]

The one thing I would say is that my biggest fear when I get behind the wheel is from other drivers. People drive overly aggressively, or completely distracted and cause significant higher numbers of accidents and traffic fatalities than do bad conditions. And situations like road construction where the road narrows by a lane or two could be handled significantly better in situations where vehicles were able to maintain speed and still merge together (which definitely seldom happens with humans at the helm).

This.

[onlineflyer]

I know there is a lot of fear out there about government control and hackers and things like that. Also fear of how these sort of systems will react to unforeseen conditions.

The one thing I would say is that my biggest fear when I get behind the wheel is from other drivers. People drive overly aggressively, or completely distracted and cause significant higher numbers of accidents and traffic fatalities than do bad conditions. And situations like road construction where the road narrows by a lane or two could be handled significantly better in situations where vehicles were able to maintain speed and still merge together (which definitely seldom happens with humans at the helm).

Several years ago I was driving on an interstate with my cruise control set at a leisurely 66 mph, since no one else was on the road, or so I thought. Suddenly I heard a bang and my van began swerving erratically down the highway. I immediately took action to control the vehicle and pulled to the side of the road. My initial thought was that something fell from an overpass but what happened was I was rear ended by another driver. The impact was so great it totaled my van. It was a bright sunny day, the other driver was obviously speeding and distracted by bickering kids.

[interested_observer]

.....
I would not get too hung up on hacking. 10 years ago, governments were hacking each other's industrial processes. We have not, as result, shut down our computer networks. It's not going to be an impediment to development. A minor irritation perhaps. But hackers and security experts will continue to square off for years to come.

... and yes, that is probably what will happen. It usually goes that way. It does not make it right. You write "It's not going to be an impediment to development. A minor irritation perhaps." ... and that's exactly what is occurring. Here is a link to an article...

• Exec accused of stealing Waymo?s trade secrets starts new self-driving company | Ars Technica

There are a couple of very telling sentences there -

In another email, he wrote that "the team is not moving fast enough due to a combination of risk aversion and lack of urgency."

“I think it’s fair to say we had different points of view on safety,”

Here in Arizona - a couple of months ago in Tempe, one of Uber's cars ran over a lady that it didn't "see". It turns out that the development engineers disabled the automagic brake system - as it was apparently somewhat annoying - if I remember correctly. So, there is an urgency to get these things out on the street. There will be causalities (or collateral damages) - too bad for the lady, however her relatives will probably hit the lottery after the court case. Kind of reminds me of the Ford Pinto fiasco. Crispy critter a few folks and life goes on (the cost benefit analysis only shows it will cost us $xxxxxx - which is substantially less than installing a shim under the gas tank). Too bad if you are involved, though. ... and what of Uber? Life was too hot to handle here in Phoenix, so they pitched their tent - fired everyone, and moved to a city back east - where the political temperature was more to their liking.

I know there is a lot of fear out there about government control and hackers and things like that. Also fear of how these sort of systems will react to unforeseen conditions.

The one thing I would say is that my biggest fear when I get behind the wheel is from other drivers. People drive overly aggressively, or completely distracted and cause significant higher numbers of accidents and traffic fatalities than do bad conditions. And situations like road construction where the road narrows by a lane or two could be handled significantly better in situations where vehicles were able to maintain speed and still merge together (which definitely seldom happens with humans at the helm).

The past has shown that this is indeed the mine set. When more could easily be done, security concerns was at the bottom of the list and we need to get everything else developed and glued together in integration, so that we can test (and then go into production). Just look at the current IoT - the Internet of Things. Autonomous vehicles are part of this, but exactly the same mistakes are being made over again, in its conception, development and early deployment. There was a recent hack of web connected cameras (embedded IP devices serving as a host) - that nearly brought down a significant part of the web. One person noticing one small design error in the hack, made one small modification, killing off the main control point - essentially fixed the problem. Otherwise, we were on the way to a reasonably sized melt down.

• The 5 Worst Examples of IoT Hacking and Vulnerabilities in Recorded History | IoT For All

... overall, things will work - well, until they don't.

[ThorSanchez]

We can’t take this discussion any further because it is already ‘political’.

Sorry, my fault, I'll stay away from such discussions in the future.

[Rondec]

... and yes, that is probably what will happen. It usually goes that way. It does not make it right. You write "It's not going to be an impediment to development. A minor irritation perhaps." ... and that's exactly what is occurring. Here is a link to an article...

• Exec accused of stealing Waymo?s trade secrets starts new self-driving company | Ars Technica

There are a couple of very telling sentences there -

Here in Arizona - a couple of months ago in Tempe, one of Uber's cars ran over a lady that it didn't "see". It turns out that the development engineers disabled the automagic brake system - as it was apparently somewhat annoying - if I remember correctly. So, there is an urgency to get these things out on the street. There will be causalities (or collateral damages) - too bad for the lady, however her relatives will probably hit the lottery after the court case. Kind of reminds me of the Ford Pinto fiasco. Crispy critter a few folks and life goes on (the cost benefit analysis only shows it will cost us $xxxxxx - which is substantially less than installing a shim under the gas tank). Too bad if you are involved, though. ... and what of Uber? Life was too hot to handle here in Phoenix, so they pitched their tent - fired everyone, and moved to a city back east - where the political temperature was more to their liking.


The past has shown that this is indeed the mine set. When more could easily be done, security concerns was at the bottom of the list and we need to get everything else developed and glued together in integration, so that we can test (and then go into production). Just look at the current IoT - the Internet of Things. Autonomous vehicles are part of this, but exactly the same mistakes are being made over again, in its conception, development and early deployment. There was a recent hack of web connected cameras (embedded IP devices serving as a host) - that nearly brought down a significant part of the web. One person noticing one small design error in the hack, made one small modification, killing off the main control point - essentially fixed the problem. Otherwise, we were on the way to a reasonably sized melt down.

• The 5 Worst Examples of IoT Hacking and Vulnerabilities in Recorded History | IoT For All

... overall, things will work - well, until they don't.

I suppose. I don't see vehicles in the US being under "web control" any time soon, at least not on the road. In my area of the country there is no web signal or cell signal for that matter, for broad swaths of the country side. And it has low enough population density that cell companies don't seem particularly interested in fixing that.

To me, self driving vehicles are still a very long way away for the average consumer who can't afford a top end car. I drive a ten year old pickup and there are a lot more people like me than those driving the newest Tesla and will be for a very long time. At the same time, I do hope that there are defensive helps that are put into vehicles to keep them from merging into an occupied lane or accelerating towards a stopped vehicle.

[ThorSanchez]

Currently the air traffic control systems around the world are being upgraded to use a system called Automatic Dependent Surveillance (Broadcast), or ADS-B. Every airplane has (or will have) an ADS-B capable transponder that is integrated with GPS, and it "squitters" information on the aircraft's position, velocity, heading, waypoints, etc, etc once or twice a second. The other aircraft and the ground control stations receive all of these squitters and computers use them to create a comprehensive, nearly real-time picture of the airspace for hundreds of miles around. It enables things like collision aviodance, and tighter, more efficient spacing of aircraft, and helps do away with set air corridors in favor of more direct routes. And all of this information is also networked around the world, which enables things like those online flight trackers you can use to see when your friend's flight into BWI is going to land. The international civil aviation organization ICAO publishes the open standard for ADS-B implementation, and although almost any manufacturer can build a transponder they all have to adhere to the interoperability standards, along with software assurance and other saftey-of-flight standards.

This system is currently in the end stages of full implementation and it will supplant the old method of using air search radars and human air traffic controllers radioing hundreds of pilots and manually telling them where and when to go.

Something equivalent to this will eventually be rolled out for cars. Every car will have a certified system squittering out position and other dynamic data multiple times a second to all of the other cars in the area and eventually this will mean dramatically safer driving.

[monochrome]

I’m not sure I want my car squittering it’s location* several times a second to anyone who has the right receiver.

I sometimes turn off my cell phone just on principle. I refuse (so far) to buy a car with OnStar or other monitoring system.