Computer experts have figured out a way to hack a Jeep, allowing them to remotely control everything from music and windshield wipers to speed and brakes. This has led to a proposal for federal cyber-security legislation, while other people question what this means for self-driving cars.
The question of the cyber-security of automobiles is almost completely separate from self-driving cars. The self-driven cars developed by Google, Volkswagen, and other manufacturers rely on several kinds of sensors to direct their travel, including GPS, lasers, radar, infrared, and optical sensors. Of these, the only one that uses the radio spectrum is GPS, and since the cars use it only to determine a route, and not for minute-by-minute driving, I don’t think it could be vulnerable to a cyber attack.
Experts believe that hackers used Chrysler’s use of Sprint communications technologies in its cars aimed at providing auto buyers with an “in-vehicle communications system.” If so, then Sprint failed to build an adequate firewall between its communications and the car’s operating controls. Chrysler responded to the hack by recalling 1.4 million vehicles that have the Sprint system, presumably so it can somehow add such a firewall. According to Wired, the hackers that demonstrated the Sprint system’s vulnerability report that
Sudden loss of penile erection can http://www.devensec.com/rules-regs/decregs109.html cialis online pill be a scary, troubling, and downright humiliating thing. Shilajit is considered an herbal mineral because it is found that men taking anti-impotency medicine don’t suffer from hormonal heart stress. viagra no prescription mastercard A doctor shall help you to determine if you are the ideal candidate for the HIFU therapy, patients are required to viagra price use this link undergo two enemas two hours prior to the procedure. mastercard cialis online Also, such drugs can result to serious or even fatal side effects on people with certain health problems. Sprint has already fixed the problem by adding security to its telecommunications.
Sprint’s response and Chrysler’s rapid recall makes it clear that federal legislation regarding the cyber-security of motor vehicles is unnecessary, as auto makers have an incentive to do their best to protect potential car buyers from hackers. On the other hand, Congress should make cars safer by forbidding the National Highway Traffic Safety Administration from mandating that vehicle-to-vehicle communication systems be included in new cars. Such vehicle-to-vehicle (and vehicle-to-infrastructure) systems would be much easier to hack than most self-driving car technologies.
Well, if you’re wise you won’t underestimate the tenacity of hackers. Murphy’s Law, you know.
In your so-far theoretical universe of self-driving cars, the specter of potentially hundreds of millions of highly computerized targets and rolling data collection systems (inevitably, as part of the so-called “Internet of Things,” as our corporate randian overlords will demand so they can collect the “big data” to sell us even more stuff) would be irresistible click-bait for hackers.
One of the problems of self-driven cars is that unlike humans they follow every traffic law, and that is irritating to many drivers.
When Delphi took its prototype Audi robocar from San Francisco to New York in April, the car obeyed every traffic law, hewing to the speed limit even if that meant impeding the flow of traffic.
“You can imagine the reaction of the drivers around us,” Michael Pozsar, director of electronic controls at Delphi, said at a conference in Michigan last week, according to Automotive News. “Oh, boy. It’s a good thing engineers have thick skin. All kinds of indecent hand gestures were made to our drivers.”
And that indicates that a problem is brewing, argues Prof Alain Kornhauser, who directs the transportation program at Princeton University. “The shame of the driving laws is that they all sort of have a ‘wink’ associated with them,” he says. “It says 55 miles per hour, but everyone knows that you can do 9 over. If that’s the situation, why isn’t it written that way—with a speed limit at 64?”
People know when to disregard a law, but it isn’t easy to reduce such implicit knowledge to an algorithm. Even if you could, there are some laws that robots should never have to observe.
“A stop sign—rather than a ‘yield’ sign—is there to make sure people have the opportunity to look both ways and see nobody is coming,” Kornhauser says. “But with 360-degree camera coverage, lidars and radars, those automated cars know in a 20th of a second whether something is coming. Why should we require them to come to a complete stop?”
http://spectrum.ieee.org/cars-that-think/transportation/self-driving/why-automated-cars-need-new-traffic-laws
The Antiplanner wrote:
The question of the cyber-security of automobiles is almost completely separate from self-driving cars. The self-driven cars developed by Google, Volkswagen, and other manufacturers rely on several kinds of sensors to direct their travel, including GPS, lasers, radar, infrared, and optical sensors. Of these, the only one that uses the radio spectrum is GPS, and since the cars use it only to determine a route, and not for minute-by-minute driving, I don’t think it could be vulnerable to a cyber attack.
I must respectfully disagree.
Self-driving vehicles need network access. Maybe not all the time (since there are plenty of areas on the public road network in North America that have no cellular access), and maybe not wirelessly (as was the case with the Mopar product mentioned above), but they need to update their digital maps and presumably fix bugs on a regular basis.
That means that the cars must have integrated in-vehicle security to prevent or deter hacking, transmission of viruses, vehicle hijacking and the rest of it.
Our Randian overlords over at Rand Corporation last year put out a study called Autonomous Vehicle Technology which has a nice section on AV communications, including DOT’s version of communication requirements:
AVs will have to have some pretty heavy duty communications security. Speed limits and other rules of the road will change when AVs reach N percent of vehicles on the road. Stop signs and other such rules are necessary for human drivers, not robots.
Speaking of robots, have you seen any good robot sci-fi lately? I mean, “Terminator” is back … AGAIN.
Lots of sex with robots on the screens lately, whether it’s AMC’s “Humans” or Alex Garland’s brilliant “Ex Machina”.
Ultimately, what these fictions have in common with the comments on this post is the theme of human insecurity.
Though I’m not sure msetty is so much insecure as cravenly afraid that self-driving cars will obsolete his pathetic little ideological niche.
Metrosucky, in a weird sort of way I suppose it is a sort of complement that you post here in reaction to whenever I post something. But coming from a troll, I also find it TRULY pathetic. So enough of my “pearls before swine” commentary here.
As for my point–though I seriously doubt you’ll get, Metrosucky, so this is for the benefit of the reasonable people who read The Antiplanner’s blog–it is based on the still valid points made about automation in Bainbridge 1982 (Ironies of Automation), http://www.ise.ncsu.edu/nsf_itr/794B/papers/Bainbridge_1983_Automatica.pdf.
Bainbridge talked about the problems that remained in industrial process control after automation, particularly the human elements that remained. And she was talking about systems that are radically less complex than the processes that self-driving vehicles (“robocars” would have to deal with in the real world, as well as their relatively very simple operating environments, to wit:
(1) the automated processes discussed by Bainbridge were relatively simple and mostly linear in nature;
(2) the humans overseeing and controlling such automation are (hopefully) well-trained trained personnel, unlike mostly amateur drivers who would have to very quickly take over in case a self-driving vehicle screws up or can’t cope with a given situation (certainly would be the case for quite a while assuming that fully functional robocars actually ever do get “released into the wild” for everyday users);
(3) Bainbridge’s automated processes occur almost exclusively in tightly controlled environments, e.g., inside industrial buildings and fenced-off properties, e.g., about as far as the chaotic environment of roads and city streets as you can get. Even then, there are plenty of opportunities for things to go wrong, from disgruntled employees to natural disasters to rats chewing through the power cables.
(4) Something that Bainbridge didn’t have reason to discuss, because in 1982 almost no industrial processes were connected to the outside world uncontrolled by their owners: the huge and growing threat of hackers, something that The Antiplanner pooh-poohs in this post. Regardless of the level of security whether robocars are connected to the “Internet of Things” (sic) and/or talk to other robocars in the immediate vicinity, the hackers will always be like rats, mice (and trolls) and other vermin: always taking advantage of new opportunities and “holes” that they uncover.
Don’t underestimate this problem; hackers like rats and mice and the smarter trolls (not like Metrosucky, however) are constantly learning, and it is simply a matter of time until new holes are discovered, potentially causing many billions of dollars in economic damage, and massive inconvenience.
Ultimately, much of the “driver aid” technology that theoretically could support widespread use of robocars (assuming we restructure our cities and lifestyles AGAIN, like what occurred when the automobile was originally adopted from 1920-present; hey, what could possibly go wrong??!!) will stay “driver aid” technology.
This will perhaps deliver a majority of the alleged benefits of robocars to individuals and society, but never making the entire leap due to (a) humans will still want to have most of the control over the driving process, and (b) economically, socially and environmentally, the “full leap” will simply not be worth the trouble or cost. For most people in the final analysis, “…you’ll only be able to pry my steering wheel from my cold dead fingers…”
(You may find this ironic coming from a strong transit advocate, but then, only a few radicals want to abolish motor vehicles completely; folks like me just want to curb their excesses and get the urban transportation ethic changed to where automobiles are adapted to cities and towns, not cities and towns adapted to automobiles. As an ethic, this stands current U.S. practice on its head).
Adapting automobiles to cities and towns, not the other way round:
New Rules of Thumb (sic) needed: http://cityobservatory.org/our-old-planning-rules-of-thumb-are-all-thumbs/
Robocars should have to meet these too, if they ever get into the wild…
Since I’m here I’ll keep rubbing it in: the ultimate absurdity of cities and towns fully adapted to automobiles: http://www.strongtowns.org/journal/2015/8/4/suburban-engagement-photos.
It’s pretty strange to take engagement photos in the middle of the street but but other than that the settings for the “suburban engagement photos” don’t look too bad. They certainly look better than engagement photos taken in a “smart growth” development. Of course, if you tried to take photos in the street in smart growth environment you’d either be run over or cause a major traffic jam.
Builder, these are the same sanctimonious jerks who get off by laughing at Walmart customers. These pathetic old men suffering from ED need easy targets and scapegoats for their flaccidity and life failures. Urban elitism is a pathology fueled by insecurity, a deep seated need to control others, and problems getting an erect ion.
msetty wrote:
Since I’m here I’ll keep rubbing it in: the ultimate absurdity of cities and towns fully adapted to automobiles: http://www.strongtowns.org/journal/2015/8/4/suburban-engagement-photos.
Mr. Setty, you might explain to the audience here how essentially every public transportation provider in the United States (be they located in a “strong” town or a “weak” town) is even better adapted to taking or accepting subsidy dollars from the owners and operators of those nasty automobiles. In fact, the transit operators are so well-adapted that most of them would shut-down immediately for lack of operating subsidy dollars if the automobiles and trucks were to suddenly go away.
I have not even mentioned Reagan’s “transit nickle,’ which provided a steady source of diverted federal motor fuel tax dollars to public transit, since most of that “nickle” goes to capital subsidies, not to pay transit employees.
Huh. Will you look at that. As soon as he’s asked to back up his specious claims, msetty abruptly runs out of pearls and suffers a sudden downsizing in his family “heirlooms”. I suppose that shouldn’t be entirely unexpected, however.
Well, if you’re wise you won’t underestimate the tenacity of hackers. Murphy’s Law, you know.
If these “hackers” represent such a threat to drivers, then why haven’t they targeted them before? There are a few instances of hackers causing minor mischief by hacking variable message signs to input funny or lewd messages, but why nothing more sinister than that? If hackers wanted to disrupt traffic systems, they could certainly hack into signal control systems and reprogram them to intentionally malfunction and cause delays or crashes. Why aren’t we seeing this?
Oh, and Murphy’s Law is not actually a law.
Don’t underestimate this problem; hackers like rats and mice and the smarter trolls (not like Metrosucky, however) are constantly learning, and it is simply a matter of time until new holes are discovered, potentially causing many billions of dollars in economic damage, and massive inconvenience.
You’re ignoring the fact that the defenders of information systems are constantly learning as well. Whenever there is a security breach, the defenders not only have an opportunity to identify the source of the breach, but to find out more about the process by which it emerged, and thus to reduce vulnerability to that type of attack in the future. Security is a constant game of cat and mouse. That is aspect of security is no different now than it was in 1982. Dystopian futures are for sci-fi movies (and perhaps transit advocates), but reality is usually a lot less dramatic.
Since I’m here I’ll keep rubbing it in: the ultimate absurdity of cities and towns fully adapted to automobiles:
That’s your idea of “rubbing it in”? Sneering at engagement photos from suburban landscapes? I think you and the entire staff at Strong Towns could use some therapy.
I think most people, given the choice between driving and doing something else during their traveling time, will choose door #2. Driving is 99% drudgery so 99% of the travelers will opt to let a robot do the driving. At some point people won’t have a choice in most places as Mothers Against Self Driving™ will have lobbied legislators to ban self driven cars from urban areas. (Because if it saves even one life isn’t it worth it?)
Economically this will be a no-brainer and will amount to savings at least as high as the savings from the automation of data collection/processing starting fifty-some years ago. Look for trillion dollar annual savings from 1) dramatically decreased auto accidents with an attendant reduction in medical bills and insurance costs; 2) dramatically reduced capital costs and maintenance costs from the reduced total number of automobiles in use; 3) reduced energy costs.
I’d bet that $1 trillion/year savings will turn out to be low. From an environmental POV, even if AVs use IC engines, the number of cold starts—when fuel consumption is double for the first few minutes of operation—will reduce pollution. And of course ride sharing will be ubiquitous and has the potential to reduce exhaust pollution.
Increased vehicle miles traveled because car travel is so much safer and cheaper? Maybe, but I think ride sharing will even that out. Socially, the only downside I see is reduced “alone time” because the average person will be sharing a ride to and from work every day. Some people won’t like that. Most people will adapt. And I have a feeling that you’ll normally share a ride with people you’ve vetted for compatibility, much like people “friend” others on social websites today.
“That’s your idea of “rubbing it in”? Sneering at engagement photos from suburban landscapes? I think you and the entire staff at Strong Towns could use some therapy.”
Yep. Urban elitism combined with narcissistic personality disorder. A new condition for the DSM.
Gotta love the urban elitism coming from people who refuse to live in dense “strong” towns.
”
If so, then Sprint failed to build an adequate firewall between its communications and the car’s operating controls.
”
Why would this be Sprint’s fault? Whomever designed the system that controls the vehicles is responsible. They should already have things like it’s interface with an entertainment system segregated. Sure, Sprint has some culpability but at the end fo the day, Fiat is the one that’s at fault for the car being controlled.
In another arena, if hackers find a flaw with AddThis and exploit it to take control of people’s Facebook accounts, ultimately that is Facebook’s fault. AddThis has some culpability, but the FB’s API shouldn’t all AddThis to do anything more than share with the account. It shouldn’t be possible for any actions from AddThis to be doing any sort of FB account control. The same sort of thing applies with Spring and Fiat.
Ten. The companies that offer auto insurance find their business decimated, as rates fall to a tenth their levels before the advent of driverless cars. The net economic surplus to consumers proves to be enormous, redirecting billions in spending on risk mediation to more productive uses. The companies that remain in the business of underwriting automotive transportation form an oligarchy to regain pricing power. They focus their energies on insuring against the catastrophic failures of the road networks, which happen rarely.
Considering the large amount of people I see using their little portable computers – aka smart phones – while in the driver’s seat of an automobile, I find it hard to believe that most people wouldn’t rather be doing something else than driving.
The same goes for walking.