Uber has hired two hackers who successfully hacked into an automobile, specifically a 2014 Jeep Cherokee, by remotely taking over the vehicle. Fears of remotely stopping or navigating vehicles then fueled online speculation from those who are already skeptical about the imminent revolutionary technology.
And this is not just about future, completely autonomous, vehicles. Just as that 2014 jeep had wireless technology built in, so do many of today's vehicles. However, vehicle-to-vehicle technology is advancing and security measures need to be advancing alongside. Similarly, we certainly do not want a roadway or entire metro area's roadways hijacked by hackers. (Okay, despite my enthusiasm for dirverless technology, I can speculate with the best of them.)
Legitimate Reasons to "Hack"
Police see similar issues looming for law enforcement, but they differentiate good hacking and bad hacking. The bad hacking is, of course, constitutes nonconsensual invasions of privacy and interference with software, information, and equipment. But police have legitimate reasons to identify people and interfere with driverless vehicles. Police might want to track someone, stop someone from violating a protection order, keep someone on house arrest, and on and on. All of these can be legitimate law enforcement tasks - sometimes requiring first that a judge issue a warrant (though that will be left to future litigation). As with all public dollars, people can weigh in or vote on these matters.
However, private monitoring, such as Google or Apple knowing where you are and what you are doing at every moment is more insidious and not subject to the Fourth Amendment the way police conduct is. As much as I want driverless vehicles and other technologies, the fact is that we want to remain a free society and I certainly want a zone of privacy from prying eyes.
Big Brother - Protecting Pedestrians from Driverless Vehicles OR a Privacy-Violating Opportunity OR Both
I assume that everyone is in favor of increasing pedestrian safety and Google and two universities are working on better pedestrian-identifying frameworks, and, presumably, software. Those universities are the University of Maryland at College Park and the University of Texas at Austin. Turns out, there are privacy concerns here. If one can detect and identify a cell phone - and its owner - as a pedestrian for purposes that are good (preventing a vehicle/pedestrian collision), then so too for evil. One writer summarized speculation in a Computer World article (scroll down) about identifying and accessing information from private phones as one waits in airport security lines. I'm thinking that fast food outlets and maybe food trucks will predict and prepare food orders while we stand on line, so there will be not waiting when we get to the counter.
Monday, August 31, 2015
Arizona Giving a Big Thank You Hug to Uber
University of Arizona is joining the Uber team to do research on driverless optics mapping,and safety. And the governor of the state could not be happier. Gov. Ducey would like to see Tuscon's streets with driverless vehicles being tested and more people in his states employed on innovative technology projects. Governor Ducey also signed an executive order that allows for the testing and operation of driverless vehicles in Arizona, (Sources: University of Arizona and the Tucson Weekly.)
In the category of what about pedestrians? comes a story form the University of Texas about how our intersections might operate with driverless vehicles. They might be reserving, or cuing up in some way, for permission or a slot to cross an intersection - all happening effortlessly because slow human brains will not be involved. This is all hunky dory for highway management, but it leaves walkers, bikers, and people in wheelchairs dependent on technology to cross a street. Yes, they are in many places currently dependent on traffic signal technology, but that is pretty simply managed. The UT-conceived system would be highly individualized for vehicles. Would non-vehicular travelers need to wear a signal to be able to cue up to cross the street?
I am skeptical. The words used in the article all refer to vehicles. News flash: "traffic flow" also includes zero-emission modes.
Also in Texas
Google will soon be testing its fully self-driving cars - no human driver - on the streets of Austin. Stay weird. Unlike many human drivers, these vehicles can "see" a football field away.
And more on mixing modes
Traffic mixing of human-operated vehicles and driverless ones will be a reality for several years. Driverless software is "teaching" driverless vehicles to predict and, thus, effectively interact with humans. Turns out, humans are rather predictable - 92 percent of the time in how they behave on the highways in the behavior they demonstrate before changing lanes. What humans need from the driverless vehicles are visual indicators, similar to eye contact, when making decisions about navigation on the road.
In the category of what about pedestrians? comes a story form the University of Texas about how our intersections might operate with driverless vehicles. They might be reserving, or cuing up in some way, for permission or a slot to cross an intersection - all happening effortlessly because slow human brains will not be involved. This is all hunky dory for highway management, but it leaves walkers, bikers, and people in wheelchairs dependent on technology to cross a street. Yes, they are in many places currently dependent on traffic signal technology, but that is pretty simply managed. The UT-conceived system would be highly individualized for vehicles. Would non-vehicular travelers need to wear a signal to be able to cue up to cross the street?
I am skeptical. The words used in the article all refer to vehicles. News flash: "traffic flow" also includes zero-emission modes.
Also in Texas
Google will soon be testing its fully self-driving cars - no human driver - on the streets of Austin. Stay weird. Unlike many human drivers, these vehicles can "see" a football field away.
And more on mixing modes
Traffic mixing of human-operated vehicles and driverless ones will be a reality for several years. Driverless software is "teaching" driverless vehicles to predict and, thus, effectively interact with humans. Turns out, humans are rather predictable - 92 percent of the time in how they behave on the highways in the behavior they demonstrate before changing lanes. What humans need from the driverless vehicles are visual indicators, similar to eye contact, when making decisions about navigation on the road.
Friday, August 21, 2015
Big News: Non-Profit to Certify Driverless Buses and Trucks
A credible name in the driverless field has popped up with a new non-profit to capitalize on and contribute to the driverless field. Alain Kornhauser, who has written on and mused about autonomous vehicle technology for many years, is leading the effort. Kornhauser is the director of Princeton University's Autonomous Vehicle Engineering Program.
The new non-profit, the Center for Automated Road Transportation Safety, will be doing research, development, and certification of driverless trucks and buses. I'm not sure if that combination has a built -in conflict of interest, particularly because a major goal of the center will be to promote the commercialization of such vehicles. Will the center be sufficiently neutral to be trusted as a source for certification?
The non-profit is a New Jersey enterprise, with local transportation players participating. New Jersey Transit, however, is not represented.
Even so, with Kornhauser at the helm, and with this early start on a potentially key piece of the dirverless puzzle that no one else has claimed, there is a good chance that this new center will be able to be the safety maven - and chief comforter of the public - for this emerging field.
The new non-profit, the Center for Automated Road Transportation Safety, will be doing research, development, and certification of driverless trucks and buses. I'm not sure if that combination has a built -in conflict of interest, particularly because a major goal of the center will be to promote the commercialization of such vehicles. Will the center be sufficiently neutral to be trusted as a source for certification?
The non-profit is a New Jersey enterprise, with local transportation players participating. New Jersey Transit, however, is not represented.
Even so, with Kornhauser at the helm, and with this early start on a potentially key piece of the dirverless puzzle that no one else has claimed, there is a good chance that this new center will be able to be the safety maven - and chief comforter of the public - for this emerging field.
Apple Speculation and Hacking Worries
After weeks of travel and lots of work - okay, mainly incredible numbers of emails - I am finally returning to writing about driverless developments. Even in just weeks, the pace of these developments is speeding up. It is not hyerbole to say that more countries, more universities, more municipalities, and more companies are joining this competition every day.
We are at the point that there is quite a bit one can ignore, for example the patent for Ford's interior of a driverless car, which looks so similar to the Mercedes-Benz design that I wonder if Ford did not have a twelve year old trace it. Or Audi's almost-but-not-quite autonomous luxury sedan. I don't want an almost driverless vehicle, personally.
What's new the last few weeks is mostly speculation - huge speculation - about Apple and worries about hacking driverless vehicles so that they stop, start zigzagging, or go straight off cliffs.
No cute photos, just news.
Everyone speculating about Apple
Almost all of the speculation about Apple's foray into the driverless world is based on the Guardian article that reported on hard evidence of Apple's secret - or perhaps secretive is a better word - development of driverless vehicles.
Here's articles with educated speculation based on the Guardian article as well as tea leaves.
Christian Science Monitor
NextGov
University News
University of Texas is getting to work on hacking/security issues with driverless technology. I'm seeing more and more articles about how easy it is to hack into any connected device, which includes driverless technology. For some reason, this and the you-ll-be-nauseous in a driverless vehicle have caught the media's attention.
And in Canada, the University of Waterloo is getting the country's first driverless vehicle transit service. Varden Labs will operate a driverless shuttle at the university. The university president hitched a ride with a second-year engineering student.
Catching a Ride on Campus
Speaking of campus driverless shuttles, a driverless pod effort is being aimed at universities, amusement parks, and retirement communities. Seems that Auro Robotics is planning to push driverless, but avoid government regulations that control our roads. TechCrunch reports that Auro is already in agreement with Santa Clara University.
Update 9/3/15: Driverless shuttle vehicles are already at the Santa Clara campus and will be giving driverless rides this semester during a three-month pilot program.
One big plus of the Auro Robotics plan is that it will emphasize driverless vehicle safety in heavily pedestrian environments. Would be great to see that driverless networks will be safer for walking than the past century of human-controlled vehicles.
Hacking Fears Abound
From the same philosophy that brought you Murphy's law - updated for the 21st century - a common refrain, this time reiterated by National Defense Magazine: Anything that can be hacked (and that's anything) will be hacked. The attention in the last week or so has been on the easy hacking of driverless vehicles. You can add baby monitors, GPS devices, and a hundred other common devices that enable us to stay connected to personal information. It is remarkably easy to hack all of these devices and no one wants to be in a vehicle suddenly speeding, darting in and out of traffic, or careening off a cliff (that's all of the California coastline, by the way). No question that software developers involved with driverless technology must address this one; the insurance industry will make sure of that. Possibly legislators as well, but they are generally behind the eight ball, not being a class of tech-savvy, up-to-the-minute individuals.
Independence Day
Meanwhile an Indian startup has created a semi, but mainly, autonomous car for people who are blind or have serious visual impairments. The driver is the navigator and the car is designed to mimic a driving experience for the visually-impaired navigator. Now that we're on the verge of driverless and we're all going to give up the steering wheel, I'm not sure this invention will go anywhere, except, perhaps, to incrementally contribute to driverless technological development.
And - why driverless for everyone?
We humans are terrible drivers. It's an informative game.
We are at the point that there is quite a bit one can ignore, for example the patent for Ford's interior of a driverless car, which looks so similar to the Mercedes-Benz design that I wonder if Ford did not have a twelve year old trace it. Or Audi's almost-but-not-quite autonomous luxury sedan. I don't want an almost driverless vehicle, personally.
What's new the last few weeks is mostly speculation - huge speculation - about Apple and worries about hacking driverless vehicles so that they stop, start zigzagging, or go straight off cliffs.
No cute photos, just news.
Everyone speculating about Apple
Almost all of the speculation about Apple's foray into the driverless world is based on the Guardian article that reported on hard evidence of Apple's secret - or perhaps secretive is a better word - development of driverless vehicles.
Here's articles with educated speculation based on the Guardian article as well as tea leaves.
Christian Science Monitor
NextGov
University News
University of Texas is getting to work on hacking/security issues with driverless technology. I'm seeing more and more articles about how easy it is to hack into any connected device, which includes driverless technology. For some reason, this and the you-ll-be-nauseous in a driverless vehicle have caught the media's attention.
And in Canada, the University of Waterloo is getting the country's first driverless vehicle transit service. Varden Labs will operate a driverless shuttle at the university. The university president hitched a ride with a second-year engineering student.
Catching a Ride on Campus
Speaking of campus driverless shuttles, a driverless pod effort is being aimed at universities, amusement parks, and retirement communities. Seems that Auro Robotics is planning to push driverless, but avoid government regulations that control our roads. TechCrunch reports that Auro is already in agreement with Santa Clara University.
Update 9/3/15: Driverless shuttle vehicles are already at the Santa Clara campus and will be giving driverless rides this semester during a three-month pilot program.
One big plus of the Auro Robotics plan is that it will emphasize driverless vehicle safety in heavily pedestrian environments. Would be great to see that driverless networks will be safer for walking than the past century of human-controlled vehicles.
Hacking Fears Abound
From the same philosophy that brought you Murphy's law - updated for the 21st century - a common refrain, this time reiterated by National Defense Magazine: Anything that can be hacked (and that's anything) will be hacked. The attention in the last week or so has been on the easy hacking of driverless vehicles. You can add baby monitors, GPS devices, and a hundred other common devices that enable us to stay connected to personal information. It is remarkably easy to hack all of these devices and no one wants to be in a vehicle suddenly speeding, darting in and out of traffic, or careening off a cliff (that's all of the California coastline, by the way). No question that software developers involved with driverless technology must address this one; the insurance industry will make sure of that. Possibly legislators as well, but they are generally behind the eight ball, not being a class of tech-savvy, up-to-the-minute individuals.
Independence Day
Meanwhile an Indian startup has created a semi, but mainly, autonomous car for people who are blind or have serious visual impairments. The driver is the navigator and the car is designed to mimic a driving experience for the visually-impaired navigator. Now that we're on the verge of driverless and we're all going to give up the steering wheel, I'm not sure this invention will go anywhere, except, perhaps, to incrementally contribute to driverless technological development.
And - why driverless for everyone?
We humans are terrible drivers. It's an informative game.
Subscribe to:
Posts (Atom)