OT: Autonomous car legislation story from Wired Magazine
I clicked on this story at Wired.com because I saw the block M in the picture associated with it, but it has nothing to with UM. That aside, it was an interesting look at the state of rules surrounding autonomous vehicle research, testing, and liability. The Michigan legislature is looking at drastically altering rules about testing and such - opening them up significantly. So much so that companies would be allowed to move into public testing, on public roads, seemingly whenever they feel they're ready. As it read through it, I couldn't help but think that its a pattern that keeps repeating itself - Michigan (the state) gets ahead in something, sits on its own laurels, falls behind, then introduces shit-tastic legislation in a vain effort to regain the lead.
Hopefully, the federal DoT will get a framework of rules in place for states to use when creating legislation for driverless vehicles. They're working on that now and expect it this summer. Which, by Fed Gov't standards, probably means summer 2018. I really hope that Michigan waits for these, and doesn't F-up the thing. Driverless stuff isn't far away, and the Big 3 needs to be a vital part of it all. But not at the expense of stoopid legislation - which puts the public unwittingly at risk.
One other thing - I'm a big fan of putting hackers in prison forever when public safety is compromised intentionally. But to legislate away any responsibility from those who have been hacked (as one of the Michigan bills does) is a step too far. The story offers no details here, but if the bill gives blanket protection, its wrong. If you develop something, you have a responsibility to secure it as best you can. Blanket protection lets you not put any effort into securing your product, and will virtually open up every API to any hacker anywhere.
I'm also concerned about the trucking industry, but I figure one runaway autonomous semi will set that back 20 years.
Sent from MGoBlog HD for iPhone & iPad
Sent from MGoBlog HD for iPhone & iPad
So much easier!
I work for a large truck and auto manufacturer. I can tell you the number one problem trucking companies have is getting good drivers.
Can't blame the drivers, especially the Long haul guys/gals. It is a tough job. Never home, idiots cutiting you off, cops looking to ticket you, etc...
There is a huge financial incentive for trucking companies to do this. Not sure how long it will take, but we have taken a class 8 truck across the US with no driver ( I beleive there was one in the cab, but don't recall).
It also scares the hell out of some car companies because who wants to pay a premium if you are not driving?
Finally, the insurance is going to be interesting. I believe today insurance covers the driver. Going forward it will have to cover the car?
Good times.
and it's going to be a massive game changer for the industry. I think it'll happen for trucking earlier than passenger cars because of the massive financial incentive you mentioned. There's already a lot of testing, like you mentioned, here and in Europe. I bet 5ish years from having at least extended auto-pilot features legal and functioning on major expressways and 10ish years from actually having unmanned trucks on expressways.
least complex set of four-or-five sigma scenarios to navigate.
Maybe the automated truck cab picks up/drops off their trailer just off the interstate.
Human drivers will do the off-expressway driving, drop off the trailer at a dropoff point just off the freeway, hook up the auto-pilot truck and it'll merge itself onto the freeway and embark on the long, boring drive the human is asked to suffer through now. It won't even have to change a single lane. The human driver then simply goes back to his next pickup and shuttles another trailer to the freeway. It's going to allow for such huge price drops in shipping that it's going to grease the wheels (pun intended) of commerce to extent that may surpass freight rail, container shipping, etc.
I don't know how much you drive, but from what I've been observing of certain semi-drivers, autonomous trucks might be a significant improvement.
There is no doubt that autonomous cars/trucks won't be perfect. There are always contingencies that can occur that will cause an accident regardless of how advanced the system. That is not a reason to not use autonomous vehicles. If the number of accidents is reduced when compared to vehicles driven by humans, even by a small amount, that should be enough to implement the technology.
yeah but when has the michigan govt ever allowed logic such as this to trump old fashioned prejudice, except when it coincidentally enriches their favored campaign donors?
wouldn't surprise me if they blow the whole thing because insurers buy them off
My main fear with driverless cars isn't hackers or accidents. It's the government using it as an excuse to get access to all information related to our cars (speed, location, etc).
I'm not a Big Brother conspiracy theorist, but driverless cars could easily lead to some significant government intrusion into our private lives under the guise of legitimate goals ("if you're speeding, we should know"..."if a parolee is crossing state lines in violation of his parole, we should know"...etc).
Sent from MGoBlog HD for iPhone & iPad
Which is another problem with them. They'll go 55 when the speed limit says 55. Meanwhile traffic wants to actually move at 65-70, so they'll be an obstacle to drivers and cause way more traffic jams than they solve. If, that is, by insisting on going 55, they don't piss off their owners so much that the owner disables the feature or turns the car back in.
That would suck! My commute is anywhere from an hour fifteen, to two and half hours - if the open parts of that was limited to 55 my commute time would increase! Screw that!! I would love an auto-pilot, but not a governed speeds well below what traffic flows at...
This is so short-sighted.
1) As cars become autonomous, traffic will probably decrease as cars move together more efficiently
2) Speed limits will probably increase as accidents become less frequent and it's safer to drive faster.
3) You'll care much less about your commute time beause you'll just be working AS YOUR CAR DRIVES FOR YOU.
I'm always dumbfounded when people with long commutes -- who you would think would be the #1 supporters of autonomous vehicles -- are skeptical of this stuff. Think about what you could do with an extra three to four hours a day, based on your commute time!
I'm not skeptical - I want auto-pilot today! But I can't afford a new Tesla, nor would it be worth the expense for the little time I'd actually be able to use it right now. I'm all in for an autonomous highway because, yes, all cars would move together. But before we get there (it's coming), there will be a big mix of driverless and driverfull on the roads. That's where I (like the commentor I replied to earlier) worry about limiting speeds when on autopilot. But even so, it would probably be worth it so long as I don't have to be ready NOW to take over (like the current implementation Tesla uses).
Gotcha.
will actually own driverless cars - at least not in suburbs or urban or moderately densely populated areas. It'll be a huge waste of money for the regular person. There will be a continuously moving network of them that you will summon using your mobile device to take you where you want and when. This is why Uber is working on it and why GM partnered with Lyft. The vehicle and the network will go together
1) As cars become autonomous, traffic will probably decrease as cars move together more efficiently
An oft-cited but completely theoretical - and not especially well backed-up - idea.
Smartphones were introduced about ten years ago and about two-thirds of the US populace has one. This for a device that costs a couple hundred dollars - or can even be free with your plan if you don't get a brand-ass new one, and is generally replaced every couple of years.
Compare that to a thing that people tend to buy once, maybe twice a decade, costs at least $25,000 if you want any kind of new tech in it, and that hundreds of millions of people refuse to ever buy new.
In other words, if smartphone market penetration to two-thirds took a decade, it probably takes three decades to get to every car on the road. Adaptive cruise control came out about 15 years ago. It's still not offered on every new vehicle and probably won't be for another five years. Then it'll take another fifteen to get all the cars without it off the road.
Meanwhile, it would likely take at least 95% of cars on the road to be autonomous for the traffic benefits to kick in. A few human drivers messing with the computer controlled traffic flow will slow everything down. And autonomous or not, traffic is always going to mimic certain rules of fluid flow, which is to say that the more stuff that tries to go through a confined space, the harder it is to push that stuff through. When two million cars hit the road at rush hour, it won't matter if they're autonomous, they'll still slow down.
won't come from this supposed "cars moving together" theory. That's not actually the theory about how autonomous cars will improve traffic. Traffic certainly won't change if regular cars are simply replaced one for one by similarly sized self-driving cars.
What will create the impact on traffic will be self-driving cars potential to reduce the overall number of vehicles on the road, and/or reduce the size of the vehicles that transport single occupants or any number of occupants less than our current cars are built to transport.
We already have highly effective ride hailing services that do a very good job of getting people in urban areas door to door. This already reducing car ownership in urban areas as users are able to forego car ownership altogether and use these services to commute. Once those are autonomous vehicles and the labor cost is removed, the economics sway even more in favor of not owning cars in urban areas and using this network of self-driving.
And yes, traffic is still not reduced until people start pooling their commutes in these self-driving vehicles. That is what remains largely theoritical. Will people choose to share part of their ride to work with a stranger? Given that it will likely be no more expensive than current public transportation and we know that people currently use public transportation where it works reasonably well, and that this will be a more comfortable, easier, faster solution than current public transportation, it's a good bet that a lot of people will do it.
And as is the case with fluid flow and traffic flow, even small decreases in flow or in numbers of cars commuting simultaneously will have hugely positive impacts on traffic flow. Yes, still "theoretical" but every company with any stake in human mobility is racing to be able to do it, so they're betting it's pretty likely.
When there's the transition into autonomous vehicles the owner will be able to set the speed with a limit on the distance. This is how the Tesla works now. It does a better job driving on the highway than I ever could.
When cars are mostly autonomous speed limits would obviously be raised.
Like it or not, the affordable autonomous car is coming and is coming faster than I ever thought it would. It's going to be such a positive for society. We'll be much safer and spend less time in traffic. It will be the quantum leap in productivity that that the computer was.
"We'll be much safer and spend less time in traffic. It will be the quantum leap in productivity that that the computer was. "
But its such a big change, we have no idea what the negatives are. Yes, there will be big positives. No, its not all going to be positives, and we may not realize what the negatives are until years into it.
Isn't that true of basically every technological leap? There's always uncertainty. But almost always, the benefits outweigh the negatives. The same argument could have been used to oppose development of the auto itself.
But Im just trying to temper the 'THIS WILL BE AWESOME THINGS WILL BE SO MUCH BETTER ALL THE TIME'.
I'm not saying we put in laws to stop it.
You will have youre idrive/Windows 3.11/Vista moments, alot.
Whatever the negatives it will not be worse than our current system. About 35,000 people die each year in the US from traffic accidents, and significantly more get seriously injured. That is a huge loss not only in human terms but also economic terms. Combine increased safety with reduced traffic, reduced emissions, increased free time and lower costs of goods and I'm pretty enthusiastic.
One possible negative, and a very big one at that, is a major increase in suburban sprawl, which has gotten pretty bad in many places to begin with. (Check out how much of northern Georgia is now taken up by suburban Atlanta.) Cars made suburban sprawl. Throughout history, people have shown they're willing to travel about an hour and a half each day, on average, and advancements in mobility have proven not to reduce that number, but to expand distances.
all the models being tested go with the flow of traffic for this very reason. Plus, once enough cars on the road are actually autonomous (it's highly likely that on many roads and urban areas it will become illegal to actually drive your own car), then speed limits will be completely irrelevent. Speed limits only exist to attempt to reduce human error on the road and to limit damge when crashes happen (the effectiveness of accomplishing those things is questionable at best). But without the possibility of human error and with very few crashes, we'll all be traveling much faster and much safer, anyway.
I wrote a really long thing above, but the BLUF of it is: We are not going immediately from all human drivers to all autonomous cars. It will be a looooooong transition. Autonomous cars will have to exist in a human-driver world for a very long time. And that means speed limits, which in turn means car companies are going to studiously avoid purposely programming their products to commit moving violations.
Besides, speed limits are not going to disappear - autonomous cars still can't travel safely down I-94 in Detroit at 150 mph. And they'll still have to negotiate things like offramps and merges at safe speeds, so they'll slow down, and consequently so will the cars behind them, and so on.
all the cars being tested now are moving with the flow of traffic, because that's the SAFEST way to do it. The new car makers absolutely will program their cars to do the safest thing programmable ahead of sticking to strict speed limits.
And if you're one who cares how quickly you arrive at your destination, you should be trumpeting the advent of the self-driving car. The more that are on the road shuttling people around, the fewer single occupancy vehicles will need to be on the road, which solves the entire problem. Time to destination has far more to do with the number of vehicles on the road than the actual speed limit, since fewer cars means you can actually go the speed you'd like to go or you can easily pass someone going the speed limit.
So the premise of your concern - that self-driving cars will slow commutes - is misguided, whether they go above the speed limit or not. But they currently do anyway!
And wow, man. Are speed limits what keep people from careening off curvy roads and merging into oncoming traffic? Of course not. It's called the laws of physics, and those certainly won't go away. Autonomous cars will, of course, slow down for off ramps and merging into traffic, and so will you! Because that's how physics works!
all the cars being tested now are moving with the flow of traffic, because that's the SAFEST way to do it. The new car makers absolutely will program their cars to do the safest thing programmable ahead of sticking to strict speed limits.
This isn't true. Google got pulled over for doing 25 in a 30 because the car was impeding traffic. Their own press release in that case pointed out their cars are programmed to never exceed the speed limit. And I simply don't believe our litigious society, that comes slamming down on automakers every time they screw up, is an environment where an automaker wants to have their autonomous car involved in an accident where the car was speeding. If you think GM and Toyota paid dearly for their engineering mistakes, imagine if they programmed a car to literally (if technically) break the law and that car killed someone.
And if you're one who cares how quickly you arrive at your destination, you should be trumpeting the advent of the self-driving car.
Personally I like driving and would rather have an hour-long commute on open roads than half an hour in choked rush-hour traffic, actually.
The more that are on the road shuttling people around, the fewer single occupancy vehicles will need to be on the road, which solves the entire problem.
I don't think autonomous cars will make carpooling more popular, if that's what you're getting at. Just because a human isn't doing the driving, doesn't mean people will start wanting to make everyone else's stop on the way home. They want to step out their front door and into transportation that will deliver them as close as possible to work. If the bus or train does that then they'll take them. But one's own car is the best option for that. Nobody wants to stop at a bunch of other people's houses and workplaces.
Autonomous cars will, of course, slow down for off ramps and merging into traffic, and so will you! Because that's how physics works!
I don't know what you're arguing against here, but yes, they are constrained by the laws of physics and that's more or less exactly my point: there are several reasons why they won't provide every traffic benefit you hope for.
Yes, that particular Google prototype is limited to 25 mph (which is what the press release said, NOT that it is programmed to not exceed the speed limit), for a variety of reasons, part of them being safety, part being regulatory. So I stand corrected that not all the prototypes are moving with traffic, but if limiting the speeds are what needs to be done to test them and get the public comfortable with them and comply with regulations that are outdated, than that's what they have to do right now.
Google's lead engineer on the project has said that their cars will be programmed to go up 10mph above the speed limit when flow of traffic dictates because it is safer to go with the flow of traffic than significantly less.
Don't worry, wherever there is a place that exists that you can make an hour long commute with open roads, you'll still be able to do it for the rest of your lifetime. No one really cares about changing how people drive in that place, if it exists at all. This is solving an urban and suburban mobility issue where commutes are currently choked rush-hour traffic. And plenty of people would rather work / read / sleep / anything but drive in traffic.
I responded to an earlier post about carpooling, and yes this is one question that remains about whether traffic will actually be improved. There are LOTS of people that would carpool if they didn't have to walk to and from a pick up stop and didn't have to park at their destination even if they had to drop a person or two off. But yes, how many people is still unknown.
And there's another factor as well. Currently, we typically own one vehicle per driver and it's a four to five person car with a trunk and needs to be flexible because it would be too expensive to own seperate vehicles for every use case. So we end up using excessively large vehicles much of the time. Autonomous single passenger cars would make commuting in a vehicle that picks us up and drops us off alone so economical, that even if people choose to own a weekend family vehicle, many people will choose the car that picks them up and drops them off for $2 without needing to park for their daily commute. Cut a current lane in half and these single occupant cars could more than double the amount of traffic moving in that lane. Huge improvements to flow without even needing to carpool.
None of this is to get into the amount of space saved by having far fewer cars parked on the side of the street (or circumventing the need to do it at all). Two lane roads could go to four.
I'm not sure what you think I hope for. I'm not expecting teleportation, but this technology will deliver safer, more productive, and more timely commutes / trips for those who want them. You are probably correct that it's going to be a gradual transition and slower than a lot of people hope and expect. But it's going to happen.
This will turn into a total effed up disaster. Why? Because the Federal Government is involved.
so they'll be able to ration that. only so many miles a year, too much environmental/highway/construction/insurance, etc. problems and they will tell you what is best for you. no problem then, right?
Anything that imnvolves the Federal Government is a disaster.
Says the person commenting on the internet...
Are you kidding me? The internet is pretty much the least regulated, wildest, freest place on the planet.
1) Here in (one of) the least regulated, wildest, freest countries on the planet, our internet is allowed to mostly reflect our societal values, yes. There is massive censorship, restriction, monitoring, etc. of the internet in other countries because it's actually very easy to do. So the internet itself isn't inherently free or wild, and the exact same big brother fears were a part of early internet discussions about how to regulate it here in this country (and still are) because literally everything done on the internet is "on the grid".
2) Since it is so easily monitored, some would argue that government programs that do monitor a lot more internet activity than most people realize, are not free enough. Queue the Snowden debates.
But most people would agree that we have the "right" amount of regulation around these freedoms, because our democratic republic was set up well to give most people what they mostly want. And I'm confident we'll do that with our autonomously driving cars as well.
Just to be clear, I would only reluctantly place the US among the freest countries in the world. Heritage has them 12th in economic freedom for 2016, for example.
And I agree with you that our system was set up well, and it does mostly give most people what they want. That last part is not necessarily a good thing, though.
And it was created as a Federal Government project
The stuff you're using and enjoying now wasn't.
Sent from MGoBlog HD for iPhone & iPad
Fair point, but they can get all of that telemetry today. Every car has all of the capability built in, but only certain features are turned on, or unlocked, through the various pricey options in each base model. But the safety telemetry is all there - speed, braking, turning percentages, various rates of motion, etc... Not quite to the 'black box' level, but if the car companies - well, insurance companies - could get it passed, we'd all have our own personal black boxes for cops (and other investigators) to access when needed.
I choose to drive a '95 whatever.
convertible in good condition. you'll be able to sneak out on the road and exceed your federally mandated mileage limitation.
You know me too well, that's just what I was thinking of.
That (the miata) is the main argument I use against my brother. One day I took a picture of the sunset over the bay while I sat at a light. Later when I sent it to him I captioned it 'The horrors of driving myself'.
He's of the opinion that when enough auto cars get on the road, the non auto cars will be made illegal.
insurance companies. "We notice your commute takes you through a high crime rate area..."
Wait till your travel information is a discoverable part of a lawsuit. 'You say you weren't at X, but your car says it was there....' or a divorce 'Your car was at Suzie Q's house 5 times in the past month at 2:00pm - 330pm...'
If you're worried, better check the tracking options on your phone. Its GPS keeps a continues track of where you are, where you've been, and route you took to get there. I think they only store the last X minutes of data, but I have no doubt that data could be accessed wirelessly in any number of ways. It would also be available to police, or ex-wife to be, with a warrant. I don't know if the phone companies (or manufacturer) keep the data archived somewhere, but wouldn't want to find out if I'm trying to hide something...
as much as possible.
Auto cars aside, privacy in this country is being silently beat to death in an alley way by technology and we're all generally walking by because dang is that phone cool.
when I go bang my mistress
they already have that if you're driving with your iphone with you, and the 'black box' records everything in the event of a potential crime or insurance challenge
if you want to escape 'big brother' or even the perv neighbor with a drone, better join a hermitage