Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Transportation

Senate Committee Expected To OK Autonomous Car Bills in Michigan (detroitnews.com) 121

Posted by msmash from the autonomous-future dept.
Michael Wayland, and Melissa Burden, reporting for The Detroit News: Michigan legislators could vote as early as next week on sweeping autonomous vehicle bills that would allow self-driving cars on any Michigan road without a human driver behind the wheel. The Senate's Economic Development and International Investment Committee is holding a public hearing on the bills at 10 a.m. Wednesday at Nexteer Automotive, 3900 E. Holland, in Buena Vista Township in Saginaw County. The seven-member committee is expected to send the bills to the Senate floor for a vote as early as Tuesday. If approved, the bills would need approval of the House before heading to Gov. Rick Snyder's desk. "We're very, very sure that this is going to move out of committee tomorrow," Sen. Mike Kowall, R-White Lake Township, who introduced the legislation, told The Detroit News on Tuesday. "We've aired out just about everything over the sun."
This discussion has been archived. No new comments can be posted.

Senate Committee Expected To OK Autonomous Car Bills in Michigan More

Senate Committee Expected To OK Autonomous Car Bills in Michigan

Comments Filter:

  • This is a first (Score:4, Insightful)

    by goose-incarnated ( 1145029 ) on Wednesday August 31, 2016 @05:14PM (#52805151) Journal
    The laws governing something are completed before we've even managed to invent it.

    • Re: (Score:2, Informative)

      by Anonymous Coward

      Michigan hosts the automotive testing grounds of several companies. In southeast Michigan seeing experimental, preproduction vehicles on public roads is common.

      This legislation isn't about change or something new. This legislation is required to maintain the status quo.

    • Re: (Score:1)

      by Anonymous Coward

      They need to ramp up the TSA so they are able to handle the extra load of screening passengers before they enter self driving cars

    • Why bring a product to market that you can't sell because it's illegal? First you have to buy the laws you need. Worry about the product later. I doubt it's coincidence this is in Michigan.

      • Michigander reporting in.
        Auto manufacturing is leaving the state, and has been for a long time. (Ever see photos of Detroit?)
        The "Big 3" aren't even domestic brands anymore, and non-Detroit brands like Tesla are threatening to take market-share once model 3 start rolling out.

        I'd wager this is out of fear. Michigan is a has-been in the auto world, and we need to get back into the game. I like this proposal.

    • The laws governing something are completed before we've even managed to invent it.

      Except that self-driving cars have been around for several years. They certainly aren't perfect, but they already have a track record better than human drivers (which is not a difficult criteria). They are ready to replace human drivers on public roads for many tasks, including routine driving on known routes.

      Will SDCs be in accidents and even kill some people? Very likely. Would even more people die if the same cars had human drivers? Even more likely.

      • Except that self-driving cars have been around for several years.

        Not on public roads in many places. Until recently it would have been illegal to use a self driving car on a public road in my state.

        They certainly aren't perfect, but they already have a track record better than human drivers (which is not a difficult criteria).

        That is debatable. A) The data comes largely from private sources with a vested interest in positive results and B) since the vast majority of "self driving" cars still have a person sitting at the wheel who is actually paying attention (because it is their job) it is unclear what the source of their advantage in safety is if any. While I would agree with the assertion that

        • I can't think of a state where it's ever been illegal to put a self-driving car on the road.

          The technology to do so hasn't been ready until recently. Why would a state outlaw what isn't possible?

          And that which is not illegal is, by default, legal.

          • Up until a few years ago, it was assumed that cars had drivers, and there might well be laws referring to "the driver" in ways that would require one to be present.

            • While that's not impossible, there are 50 states and I haven't read all of the applicable laws in all of them, I've never seen a case where that's true. The laws I've seen place conditions on the driver, but do not address one not being present, making a self-driving car de facto legal.

    • Yeah, I noticed that too. Shouldn't this technology have moved farther BEFORE legalizing it?

  • Can't wait to see these things navigate black ice.

    • Re:They will be great on icy roads (Score:5, Interesting)

      by 0100010001010011 ( 652467 ) on Wednesday August 31, 2016 @05:22PM (#52805203)

      Thermal sensors will know the temperature of all sections road within stopping distance. Cameras that can see stuff outside of the visible band of radiation will see the difference in reflection between ice and normal road. A 360 deg continuously LIDAR/RADAR will know the position and velocity of everything within stopping distance.

      It won't be checking its texts, facebook, looking in the rearview mirror for 1/2 a second. "Black Ice" won't confuse it because it happens to be the same color as the road surface.

      Yeah, I can't wait to see how they'll do. I really can't wait for the autonomous vehicle rally races.

      • Um knowing the temp of the road isn't gonna really help much less it assumes its black ice all the time, throw on to issue of when roads are COVERED in snow so you can't see lines to know where the road is, and no GPS isn't gonna help.

        • There probably is a way to determine the difference in road surface. Even if it's not by seeing it ahead of time but just correctly responding when the vehicle hits the black ice patch. Good modeling of the vehicle and modeling of different alternative actions could easily result in better control on such surfaces than human drivers can achieve.

          • Its called black ice for a reason.

            • Re:They will be great on icy roads (Score:4, Informative)

              by 0100010001010011 ( 652467 ) on Wednesday August 31, 2016 @06:08PM (#52805419)

              Because it's invisible to the human eye? A poorly designed sensor that only detect stuff on the EM spectrum between 400-700 nm wavelength? (And in some people it can't even differentiate between red and green).

              What does black ice look like on the IR spectrum? UV?

              The more I read slashdotters comment on modern technology that some of us have experience with the less and less confidence that I have any of them know anything.

              throw on to issue of when roads are COVERED in snow so you can't see lines to know where the road is, and no GPS isn't gonna help.

              So how does a human stay on the road? We somehow manage with two (or one) optical sensor with limited range of view driven through a neural net that has limited bandwidth and slow propagation delay between nodes. In older cars without ABS or power steering you could sometimes add a 'touch' sensor to that because you could get feedback as to what was going on through the brakes or steering wheel.

          • They're going to have to model the shape of the road.. where I am they don't plow down to pavement everywhere. If the tires don't match the ruts you can easily spin the car sideways and slide uncontrollably. The alternative is to drive 5 mph everywhere, which is what I am very afraid these vehicles will end up doing.

            • Because a human has a model of the shape of the road? How does the human get into the ruts? Magic?

              Why is the sideways slide 'uncontrollable'? Because *you* can't control it?

              • It's uncontrollable because the movement of the car is perpendicular to the direction of the wheels.

                • And how did it get there? Sounds like the feedback controller manipulating the actuators was tuned wrong, didn't have the right sample rate or had a bad sensor (perhaps it was checking a text message).

            • This is in fact how they already do it to an extent. Nearly all the autonomous vehicles in testing have lidar sensors which do give that kind of shape for modeling. They've got some really cool, lower cost LIDAR sensors like these. http://spectrum.ieee.org/cars-... [ieee.org]

              As a side note, the actual hardware is doable. 6 or so LIDAR sensors, a dozen or so fixed position camera, a couple of radars, and a computational architecture that probably uses FPGAs and custom ASICS for the machine vision and machine learni

              • Do you know how much seat belts cost? $5. Car companies had to be forced by law to put them in.

                • Not comparable. The seat belt doesn't increase the value of the car to the buyer, at least in the early days when the general public didn't like to wear them. If your car's autonomous, it's got concrete financial value.

                  1. You can live farther away and save money on rent and mortgage, doing your first hour or 2 of work as the car drives itself
                  2. You can save money on gasoline because while you're at work it can go to a charging station and spend a couple hours charging at a robotic charging station
                  3. Yo

                  • 1. Won't matter to most people. I don't live an hour for work because I have to come home, have supper with my family, and then take my kids to extra curricular activities. Time away from home is time away from home, doesn't matter if it is more productive time.
                    2. How much will that charging station cost? Parking where I live is now $18 a day with no electricity. People will smell profit potential.
                    3. You're assuming an automated car will just be able to find a spot on the street somewhere, as opposed

                    • Well, (1 and 2) are specific to your situation, not most people's situations. (3) is more a "by definition" thing - if you can't trust the car to find a spot to park in an ordinary lot somewhere and not hit any people or other cars in doing so, it's not really a smart enough vehicle to be considered fully automated. (4) is possible to deal with, it's not the same as public transit because people using your car as a taxi are not anonymous. Also, it could make a stop at a cleaning place before returning to

                    • All I know is that there are a great many automated features in luxury vehicles that are used to sell luxury vehicles. For example, adaptive headlights have been around for at least 10 years, and still only put into luxury vehicles. Because that is how capitalism works; if they find a feature that sells cars they will use it to sell the most expensive cars. They don't care about the lives that they would save if the technology was accessible to everyone. Automation will not every be available to everyon

                    • Ok, you've changed your argument. I concede this is a massive social problem because the truth is, if several million drivers lose their jobs rapidly over a few years as automation becomes huge (I'd expect it to take about as long as it took for iphone like smartphones to become the overwhelming majority of phones sold - 5 to 10 years, once automation makes economic sense).

                      But it's not one you can claim won't happen. The costs of the physical hardware in each vehicle won't stop it from taking over. How m

        • Re: (Score:2)

          by dave420 ( 699308 )

          If a person can deal with it, a computer with better processing and hardware will also be able to. How do you figure out where the lanes are? Why do you assume it's impossible for code to do that too?

        • One of the scariest experiences I ever had driving involved black ice in a heavy snowstorm. I couldn't tell it was there until I started to lose control. I quickly regained it and then lost it again a moment later causing me to do a near 360 and for a moment I was sure the cars behind me were going to slam into me or that I was going to end up in a ditch. Instead I came to a near perfect stop at the red light. In retrospect I should not have even tried to stop for it.

          Worst case scenario: It would

      • What vehicle has this technology in it? Google cars? A lot of people talk about 'what autonomous cars can do' and then list technology that no company has made work in a car ever. The fact that the technology exists and could theoretically be used in a car doesn't mean a company will make it commercially viable and reliable enough to put in an actual vehicle.

        • What computer has the technology not to use punch cards? IBMs? A lot of people talk about "compilers" and "programmers writing their own code" and then list technology that no company has made work in a car ever. The fact that punch cards are easy to use and we already have operators trained to use them means there exists a company to make it commercially viable and reliable enough to ever let a programmer type Fortran.

          Additionally, how do you know what autonomous cars can an can't do? Are you relying on pr

          • Ok well then let's stop talking about automated cars and look forward to transporter technology then, if our conversation is to have no basis in reality.

            • Other than the fact that companies had prototypes back in 2004. [wikimedia.org] And they've come a long way in 10 years. Another 10 years now that it's near commercialized and out of research think tanks and by 2024 there are going to be multiple on the road offered by multiple companies.

              • Does your crystal ball mention if anyone will be able to afford one of these vehicles in their driveway?

                • Does your crystal ball mention if anyone will be able to afford one of these vehicles in their driveway?

                  Tesla cars already come with all the hardware needed for full self-driving. They just need a software upgrade. Some people can afford them.

                  The cameras are very cheap, as are the ultra-sound sensors. Automative radar is affordable and dropping in price. LIDAR is more expensive, but many cars, including Tesla, don't use LIDAR. The full package of sensors and computer should only add a few hundred to the cost of a car, and you are likely to quickly recoupe that with lower insurance premiums.

                  • Tesla cars don't detect the contour of the road, to my knowledge and can't see up to the top of the car. Lidar isn't used because it can't see through fog and gets fooled by simple things like a bag blowing in the wind in front of the car.

                    • You're right. I concede. You heard it here companies working on autonomous vehicles, fluffemutter said that it can't be done. Might as well give up your R&D departments now.

                    • I didn't say it couldn't be done. I just said that it is a far stretch to assume it can be done. So far of a stretch that we can't really talk about it as if it is a viable possibility. I would concede that *if* thermal sensors are practical in cars some day then they may be used for this purpose but no one has any way of knowing that at present time.

                    • Re: (Score:2)

                      by dave420 ( 699308 )

                      It's not a far stretch at all - they only have to be as good as humans. If people can navigate this stuff, then autonomous cars can. Their improved feedback from the wheels would mean they can detect dangerous driving conditions better than humans. They don't have to be perfect, and they will improve all the time. You have to show why there is some impenetrable barrier for these cars which humans can comfortably sail through. Until you can do that, there is no reason to think the cars won't be at least

                    • Impenetrable barrier? How much of real life can current AI really understand? It can win at Go.. Bid deal. A bunch of well known calculations stacked up on top of one another billions of times. You aren't fully appreciating how tuned the human mind is to anticipating events in the real world and I can't explain it to you.

                    • > A bunch of well known calculations stacked up on top of one another billions of times.

                      What exactly do you think "anticipating events" is?

                      Do you think a auto driving car is going to 'forget' how it is to drive on snow and ice every fall? Is an auto driving car going to 'forget' that deer are more active during a certain time of year and not anticpate them?

                      It doesn't need to 'anticipate' an event because it's already calculated all outcomes and chooses the best which is what it does in Go.

                    • You can't calculate when deer are going to show up. An automated car will have to have some sort of long range heat sensor for deer and other wildlife, good to 100ft or so. Unless you want the car to drive slowly forever more just because deer are prevalent at that time. Can current LIDAR even tell the difference between a bag blowing in the wind and a rock of the same size being dropped from a bridge? Will it detect a person on that bridge that might drop a rock?

                    • > You can't calculate when deer are going to show up

                      Um, you can do exactly that. It's what actuaries for insurance companies do for a living. How often a person collides with a deer is something that is both measurable and calculable

                      > An automated car will have to have some sort of long range heat sensor for deer and other wildlife, good to 100ft or so.

                      Uh, ok. FLIR [flir.com] has existed for years. It's cheap enough now they're putting it in cell phones.

                      > Unless you want the car to drive slowly forever more

                    • Your stance would make sense if anyone has made this stuff work. But as it sits right now, apparently Tesla Model S's will drive into a trailer. If they will do that, why would I think they would avoid a falling rock? I'm just calling things as I see them. I don't really care if the technology is out there, the fact is that autonomous vehicles don't deal with these things now and may not for a great many years to come. Especially at a consumer level.

                    • Your stance would make sense if anyone has made this stuff work.

                      They have. Just because you don't know they have doesn't mean it doesn't exist.

                      apparently

                      You know this from analyzing the Tesla data? You've run the data analytics? Tesla is a very new company to this game as well and rushed to be the firsh to market.

                      Especially at a consumer level.

                      What ever you want to believe.

                    • Tesla data only shows us how safe Autopilot is in the circumstances where Autopilot works. You can't compare that to a human right now. Turn autopilot on in your driveway and ask it to to drive all the way to your destination without turning off (like a human has to) and then see how safe it is. Just as an example, how many accidents happen on the highway just because a human tried to pass another vehicle and misjudged the distance they had? Autopilot doesn't pass other vehicles so you can't look at it

                    • Highway fatality statistics for humans are much, much worse than those of computers.

                      Just as an example, how many accidents happen on the highway just because a human tried to pass another vehicle and misjudged the distance they had

                      Exactly. A computer would know exactly how far another vehicle was. It would know in less than a second how fast it was going, has calculated how long it will take to pass, knows the vehicle weight and exactly how fast it can accelerate. Before a human would be through questioning if they should pass the computer would have already made a better decision.

                      Why do you continually bring up Tesla? They're the last ones to the gam

                    • So which of those companies are performing safe passing of other vehicles at highway speeds with technology that will be widely affordable in the near future?

                    • Audi: https://www.youtube.com/watch?... [youtube.com]

                      Georgia Tech: http://www.popsci.com/georgia-... [popsci.com]

                      Oshkosh: https://oshkoshdefense.com/tec... [oshkoshdefense.com]

                    • You realize autonomous off-road vehicles are way easier to design than civilian road vehicles.. right?

        • Most things aren't commercially viable when invented. Most products have gone through phases like "This would be cool"..."this mostly works in the lab"..."we could probably sell this if we could get the price down"..."we've sold a few in test markets and these are the problems we're finding"..."this works well enough but it's still too expensive for large-scale merchandizing"..."this is ready for prime time"..."BUY IT USED AT JOE'S".

          • Fine, but people shouldn't be talking about it like it will happen because it might not.

            • Plenty of people are working on it, and every part we need is clearly possible. It's a matter of putting things together and refining them.

              What we will not have any time soon is highly intelligent obstacle prediction. The software can notice a ball coming into the street, and anticipate a child following. It can note whether cars along the street are ready to pull out or not. However, it won't be able to do what I did once: there was a pedestrian on a corner when I approached an intersection, and I

              • You're not sure if it is at all important? Well it sure is if we don't want cars to be hitting pedestrians! Automated cars will need to anticipate anything and everything a human can or the lawsuits against the companies will be plentiful and will make them financially impossible. If an automated car hits that kid that darted out onto the street, that kid's parents should absolutely sue the automated car company or at least get a big insurance settlement. I can't even think of all the edge cases that ha
      • Comment removed based on user account deletion

        • Why not both? Split the sport into 2.

          Humans racing in Group B.

          Robots racing in the "battle bots" series. Screw the 110 kg (~250 lb) weight limit on battle bots. I want robotic warfare on a racetrack sized arena.

    • They'll probably be programmed to not slam on the brakes, so it will be an improvement over their human counterparts.

    • Can't wait to see these things navigate black ice.

      During snowstorms last winter, Tesla recommended that drivers engage Autopilot because it could handle the icy roads more safely than most humans.

      It is amusing that when people want to point out a limitation of SDCs, they often pick something that SDCs do particularly well. SDCs handle low traction situations better than humans do.

      • Doesn't Autopilot need to see lane markings?

        • Doesn't Autopilot need to see lane markings?

          No. Tesla collects location information from its cars. It then takes that information, throws out any outliers, and averages it to get the center point for the lane. With that information, it can navigate using GPS only, with no lane markings.

    • I live in Minnesota, and I have problems when I encounter black ice. It wouldn't take that much for a self-driving car to be better at handling it than I am.

  • "We've aired out just about everything over the sun."

    Leave it to a republican to even fuck up such an old ass saying.
    Perhaps the sun is underground where he is.

  • Still have to use a dealer (Score:4, Insightful)

    by sjbe ( 173966 ) on Wednesday August 31, 2016 @05:52PM (#52805341)

    Great. But we still can't buy a car without involving a greedy pointless middleman.

    • Michigan being protectionist in favor of old established companies?
      Shocked, I tell you. Shocked.

      This will change once the sales tax folks start noticing how many of us are hopping over to Ohio or Indiana to buy Tesla's direct.
      Takes money to fight money.

      • Michigan being protectionist in favor of old established companies?

        Go figure. What I don't really get is why Ford, GM and FCA aren't making a bigger stink about getting rid of the dealer networks. I realize there is a lot of short term pain for them to do that but the dealers really provide them no benefit at all. In fact they really hurt their business. The sell marginally less cars because of the middleman markup, nobody trusts the dealers, the dealers capture a ton of repair business, etc. I don't get why they aren't maneuvering to buy out or otherwise get rid of t

  • Michigan lawmakers still think their state is relevant in the high-tech auto industry.

    • Re: (Score:2)

      by HiThere ( 15173 )

      Well, this will make it relevant. Driverless cars need to be tested somewhere, and Nevada ghost towns aren't really sufficient. Perhaps they'll decide to make them near where they can test them.

    • Totally agree. BUT - if this passes, it's a great proving grounds for winter driving. Nevada has hot & dry down pat. Cars need to be tuned for driving in sub-optimal weather, and man, do we ever have that.

      Drop an autonomous car in the lake-effect snow region, and that problem will get solved real quick.

  • Self-driving cars are still vaporware. If the idiots in Detroit put cars on the road that are not capable of perceiving ALL road variables, and thus cause deaths, the legislators who pass this bill; the Governor, and all senior automotive execs should be help *criminally* responsible.

    • US law indemnifies law makers and other government officials in these cases.

    • criminal cases don't have EULA's or NDA's to block from getting out. And I want a hard ass Judge like the one from My Cousin Vinny to hold people in contempt of court for trying to pull the EULA's, or NDA's BS to get out of giving out logs / source code. Just wait for an deadly accident and a criminal case to be opened.

    • Agreed. Self driving cars/cabs dont currently exist in Singapore. Like not at all. Completely vaporware.

    • Wow, a straw man, true Irishman, and ad hominem all in one post. Good job!

  • Now the people in Flint can have bottled water sent to their door by a driverless vehicle. How about cleaning up the plumbing in Flint? Maybe we need a plumberless plumber to do the work.

  • 2016 update on Google Self driving cars. A much watch video if you are interested in this stuff. https://www.youtube.com/watch?v=Uj-rK8V-rik

  • Do they have to certify the auto driving system? What is the criteria? Can I just slap in my own homemade system in an existing car like George Hotz and just unleash it on the streets?

Slashdot Top Deals

You know you've landed gear-up when it takes full power to taxi.

Close