5 ms·
Ask HN: Why isn't there more effort to build highway-bound self-driving cars?
Why isn't there more effort to build highway-bound self-driving cars?
1. Most commuting is on highways. A highway is a controlled environment with very few edge cases--less than existing L4 areas like Arizona and Pittsburgh suburbs.
2. Highway driving captures 90% of the value of a self-driving car, certainly for commuters. Any long trip, where it is valuable to disengage, spends most of its time on the highway.
3. An autonomous highway in a single city would save its commuters an absurd quantity of time and attention.
4. Most new cars can navigate from onramp to offramp. But L4 providers haven't taken the step to let drivers disengage, despite supporting L4 driving in limited non-highway environments.
5. One might claim that it isn't as valuable for ridesharing, but it's is hard to believe that there isn't enough value for commuters, or there isn't enough funding.
6. To handle edge cases, it is always legal to slow, even stop on a highway, and there is always a shoulder to pull over.[1]
7. For rare edge cases, a remote driver can serve as a backstop, as in Waymo's L4 driving.
8. Imagine how much better highway traffic would be if one could disengage and read, work, or think.
I understand that it involves government collaboration, but companies have created significant regulatory change when necessary. For some highways, it requires no government collaboration.
[1] There are rare exceptions, but the geofence could extend only to highways with a shoulder.
- rthomas6 5y agoHow is this different than a bus? This is what buses do right now.
- flerovium 5y ago1. Buses make frequent stops and are much slower, especially in the US. 2. Most Americans commute from suburbs to cities. This requires a few minutes to get to the highway from a suburb, which is spread out too much for a bus to be practical.
- zamadatix 5y agoI'd pick high speed rail before this. Not only do I think it's better for the use case but I don't think anywhere near 90% of the value of self driving cars is on the highway. If anything this would probably increase car traffic delays as cities would become even more inundated with cars once you got off the highway.
- muzani 5y agolaughs in southeast Asian I've literally quit jobs because the 3 hour daily commute is too much, and I'd rather not go the 5 hour route by public transport.
- flerovium 5y agoThat might be even better, but a single company can make self-driving cars without reforming the American suburban model. Practically, even in cities with excellent commuter rail (e.g. Boston), car commuters still outnumber rail.
- zamadatix 5y agoIt's not the suburban model that would need to be reformed, cars (self driving or not) are fine out there, it's about reforming using cars to get into and around cities. You don't need a train station 1,000ft from your house you just need rail fast enough to make it worth driving your car to the train station instead of driving it straight in. MTBA in Boston doesn't provide that as they aren't high speed, though they do have uniquely decent coverage for an American city. No matter how low effort you make the middle part of the journey it's not going to make much of a dent in the amount of traffic one has to deal with as there is no way to get enough cars through a city to meet the demand if it were quick and easy. Now if this were about full self driving start to finish I could buy it but just making the highway portion better doesn't really fix anything at scale except for long distance truckers and family vacations.
- tennien 5y agoThat’s interesting: I’d think a 30min commute would be much better if the 20-25min could be spent disengaged.
- zamadatix 5y agoSo do I. I just don't think the portion outside of the disengaged part of the commute would stay the same. Same as every other time car commutes to cities were made easier so more cars came in slowing them back down to the point it's too inconvenient. "If anything this would probably increase car traffic delays as cities would become even more inundated with cars once you got off the highway." Even tripling car capacity wouldn't oversaturate induced demand which is why I say I'd take high speed rail which can actually oversaturate induced demand due to its speed and density. Or full self driving where there is no portion where you can waste time due to increased congestion.
- guilhemherail 5y agoThe risk for fatalities goes up with speed. When a vehicle is driven at ~65mph on the freeway, you have less margin for error than when a vehicle is driven in a city at ~20mph. If something comes up, a remote operator may not necessarily have time to safely manage the situation, and the risk of having a catastrophic event is as high as it gets.
- tennien 5y agoMost fatalities happen off highways, despite highways accounting for a disproportionate number of miles driven. Accidents still happen, but it’s hard to believe they would be less safe than human drivers, when most accidents are caused by fatigue, reckless driving, illegal maneuvers, and unusual driving on the part of the operator. It’s even harder to believe that it wouldn’t become safer with effort, when both the road and the behavior of cars is far more predictable.
- guilhemherail 5y agoThe fatalities with AVs would most likely not be distributed in the same way as the fatalities with human drivers: Most fatalities, when humans are driving, happen because of speeding (ex: 45 mph in a 25mph zone), drunk driving, fatigue/distraction, etc. Most fatalities, with AV, will come configurations where you cannot beat the physics and stop, even with superhuman reflexes. For me, this means situations with higher speeds. The system will have less time to compute everything, and if there is an impact, the risk for fatalities is far higher: https://nacto.org/docs/usdg/relationship_between_speed_risk_fatal_injury_pedestrians_and_car_occupants_richards.pdf https://nacto.org/docs/usdg/relationship_between_speed_risk_... At slower speeds, you can build a system with a defensive driving style that mostly eliminates by design the configuration where you could be the cause for a fatality. As a manufacturer of AV, you would probably want to start deploying your system where it is the least likely to fail / cause deaths. Also, as you said (and it is probably true), most fatalities happen off highways despite the disproportionate number of miles, which means human driver are actually pretty good with highway safety, making it harder for an AV to be better than a human.
- tuatoru 5y ago1. Commuters' start and end points are not on highways. The value of commuting is in arriving at your destination. 2. Attention. How does a car ensure that its driver re-engages when required? Imagine what would happen if thirty or fifty cars pull over because their drivers are busy finishing emails. 2A. If self-driving cars become common on a given highway, how do they cope with griefers? (Other drivers who manipulate the software into giving up, or worse.) 7. Having remote drivers is about as expensive as having an on-board driver. 9. Safety. Self driving software is not infallible, and there are other vehicles on the highway, and unusual situations can develop. 9A. Kinetic energy. Major accidents at highway speeds can be lethal. 10. Publicity. Accidents on highways can also be spectacular/photogenic. 11. Liability risk. Anything that can be blamed on the software, will be.
- tennien 5y agoMore fatalities happen off highways, where roads are in unusual condition and there are more edge cases. 1. Wouldn’t a 30min commute be 1/3 as worse there’s a 20-25min section you can disengage? 2. It can force reengagement via the same mechanism for L2 highway driving, but there is a period where one need not pay attention. There are also existing car lots on the sides of highways for maintenance, and they could be added and expanded. At worst, the shoulder provides space for hundreds of cars per km. 3. greifers are a problem for any self-driving system 7. The remote drivers need engage only in exceptional circumstances, as the driver would. 9-11. apply to existing geofences, but with far more added complication, including pedestrians and cyclists.
- SkyPuncher 5y agoI have a Hyundai Palisade with L2 driving. Basically every new car coming out has the same features. L2 handles 98% of highway driving needs. I've driven 500+ miles including through Chicago rush hour traffic only needing to handle lane changes, interchange merges, and highway exits. Sure it's not set-n-forget, but it's close enough. For me, it's literally been life changing. I have terrible knees from sports injuries. Long drives use to mess me up for days. With L2, I can cruise for days without feeling uncomfortable.