Viewing 26 posts - 81 through 106 (of 106 total)
  • Self Driving Cars?!
  • wilburt
    Free Member

    autonomous vehicle failed to recognise a cyclist.

    Not a mistake the psychopath’s they let drive First Eastern buses ever make.

    The standard of driving we accept from humans could be surpassed by the average smartphone.

    The-Swedish-Chef
    Free Member

    Some interesting thoughts in this thread.

    Initially you’ll see the development of existing driver aids and the addition of extra sensors to enable unsupervised driving within controllable environments like motorways and other pedestrian/cycle free roads. This will pave the way for standards, laws, acceptance, and further development that will lead to the type of point to point autonomous driving seen as the end goal by a lot of people.

    Cougar
    Full Member

    The standard of driving we accept from humans could be surpassed by the average smartphone.

    It really couldn’t, or we’d have done it by now.

    badllama
    Free Member

    I’d rather drive thanks

    [video]https://youtu.be/L1UxZJ9owXY[/video]

    😀

    kentishman
    Free Member

    Initially it will be good as you can do other things.
    Then the boss will see that you can now work on the move so we will all just work longer.
    Then when the car can talk to each other a class system will be introduced where if you pay more the other cars will get out of the way so if you have the money you can get across town faster. And the rest of us will be still in a jam.
    But surly the easiest things to do first are not the cars but the trams and trains.

    edhornby
    Full Member

    Given the terrible system of deregulated bus services in the UK, it wouldn’t surprise me if Waymo buses are the first commercial foray rather than the personal use, and delivery before personal also – something on the lines of overnight delivery of bogrolls along a minor motorway, small electric buses on a rural route that the likes of first/stagecoach can’t be arsed providing

    wilburt
    Free Member

    We dont do it because people like driving and other people like making money from people who like driving.

    edhornby
    Full Member

    I’m not suggesting that private cars will just disappear (the ebook hasn’t replaced the paperback, they coexist) because there are folk who have a genuine need for a car/van, a plasterer wouldn’t be a waymo user, parents of small children will have a car with childseats in and there are lots of people who have weekend cars and will carry on using them.

    but the urban commuters are the ones who will be the adopters – the idea of getting into a car waiting for you and not having to park, maybe even picking up a share passenger that is waiting on your route without deviation, is surely tempting

    Northwind
    Full Member

    I keep coming back to “I like driving”- because I do. Sometimes. But actually not that often- most of the time it’s just a means to an end and often, it’s rubbish. And you miss out on things, too- sure I love the feeling of driving up to Fort William via Glencoe but it’s generally not really that much to do with me being in charge, it’s mostly the travel. And for every bit where the driving is great- Loch Lubnaig frinstance- or where it’s the whole experience of driving and surroundings, like the road along loch laggan- there’s more where it’s not, or where I could be looking out the window, or reading a book… And that’s even leaving aside the motorway-bashing.

    So yeah, I like driving and I’d miss that. But not enough to keep doing it manually if I could switch it all off, because there’d be more than enough payback. Ideally I’d like a balance- I want to autopilot to North Ballachulish then drive the german road to Kinlochleven, or autopilot to Callendar then take 15 minutes out to batter around the Duke’s Pass and back, then let the car take over again. But if that’s not an option then I’ll take the lot.

    hairyscary
    Full Member

    Driving is a chore that takes me to places where I’ll have fun or do essential stuff. What joy it would be to key in your destination and then spend the next hour, two or six chatting, reading a book, watching a film, admiring the view or getting some sleep.
    The sooner the better.

    funkmasterp
    Full Member

    We should be working on teleportation. Balls to any other form of transport for commuting. Me, bike, top of hill, go! 🙂

    In all seriousness I don’t think I’ll see roads with majority self driving cars in my life time. It will be slow and steady and any accidents that result in loss of life will cause the tabloids to go in to overdrive, killer car headline mode I reckon. Car ownership is like house ownership in this country. It’s ingrained as something to strive for and a sign of status. Changing that ownership mindset would take a long time.

    dissonance
    Full Member

    Accidents per mile way better than human drivers.

    If that is the Tesla claims from what I recall they used a rather relaxed approach to statistics. Namely comparing the accidents per mile for the Tesla approved usage vs all human driven miles.
    Its worth noting the original providers of the Tesla tech (Mobileye who are in the process of being brought by Intel) fell out with Tesla due to, depending on who you believe, Tesla make way to big promises.

    FuzzyWuzzy
    Full Member

    I think it will be great once the kinks and legal situation is worked out, it’s just that might take a long time.
    I can see initially there will be some rule that the ‘driver’ must always be ready to manually override and take control and therefore the ‘driver’ remains the person responsible for the car. That sounds like a nightmare to me though, I’d rather actually be driving to help keep focused than just sat there stressed-out ready to grab the steering wheel and second-guessing what the car is doing all the time.
    Eventually they’ll change it and cars will stop even having a manual override/steering wheel option but I think it will take a generation or two for people to start trusting them and given all the security flaws already found in car computer systems how long before someone dies (or kills someone else) as a result of their car getting hacked? I probably sound like I need to buy some tin foil but security always comes second to the rush to get a product out there with features customers want.
    Not sure it will kill radio either, I get car sick pretty quickly if I read or watch TV/video when being driven so listening to the radio would still be my choice.

    The-Swedish-Chef
    Free Member

    I can see initially there will be some rule that the ‘driver’ must always be ready to manually override and take control and therefore the ‘driver’ remains the person responsible for the car

    This is defined as level 3 AD and is a complete legal nightmare, this is where Tesla are today.

    Level 3: Within known, limited environments (such as freeways), the driver can safely turn their attention away from driving tasks, but must still be prepared to take control when needed.

    The next level is where the industry needs to get to:
    Level 4: As level 3, but no driver attention is required. Outside the limited environment the vehicle must be able to enter a safe fallback mode – i.e. park the car – if the driver does not retake control

    e.g. having a safe stop fall back option if the driver does not take control

    Northwind
    Full Member

    FuzzyWuzzy – Member

    That sounds like a nightmare to me though, I’d rather actually be driving to help keep focused than just sat there stressed-out ready to grab the steering wheel and second-guessing what the car is doing all the time.

    Yup. I consider myself a decent, attentive, sensible driver but if I’m on a 6 hour automated drive and some driver intervention is suddenly needed after 5 hours then we are crashing 9 times out of 10, humans just aren’t wired for that.

    The mechanical failure recovery stuff has got to be a nightmare- how does the automatic car deal with a blowout, or total engine cutout in the 3rd lane of a motorway

    The-Swedish-Chef
    Free Member

    ow does the automatic car deal with a blowout, or total engine cutout in the 3rd lane of a motorway

    If its in AD mode then I’d say that it will be able to cope better than a human as it will be able to process information and react quicker, this is in comparison to having to over compensate for driver interaction in normal driving mode.

    Northwind
    Full Member

    The reason I used that example is I’ve done it (the breakdown in the 3rd lane, in heavy traffic) and it was a very human, reactive process to get across to the hard shoulder (stopping on the central reservation = bad juju). Lots of mad indicating/flashing of lights, car body language to create and maintain gaps etc, it was grim.

    In a fully automated network your internet-of-cars or swarm mechanic or whatever could take care of it but with autonomous vehicles that’s a challenge- the big difference is that it required me to get a response from other road users rather than just reacting

    richmtb
    Full Member

    Level 3: Within known, limited environments (such as freeways), the driver can safely turn their attention away from driving tasks, but must still be prepared to take control when needed.

    This is pretty much how modern passenger aircraft operate. Within the parameters of a standard flight they are capable of full autonomy – including take off and landing. Pilots need to intervene if things start to go wrong. But crucially pilots tend to be better trained than your average driver. They also tend to have a lot more time and space to deal with a problem. Sure the skies are getting busier but they aren’t the middle lane of the M25 at rush hour.

    Yup. I consider myself a decent, attentive, sensible driver but if I’m on a 6 hour automated drive and some driver intervention is suddenly needed after 5 hours then we are crashing 9 times out of 10, humans just aren’t wired for that.

    Yep very much this

    Cougar
    Full Member

    but the urban commuters are the ones who will be the adopters – the idea of getting into a car waiting for you and not having to park, maybe even picking up a share passenger that is waiting on your route without deviation, is surely tempting

    We already have this, we call them taxis.

    In a fully automated network your internet-of-cars or swarm mechanic or whatever could take care of it but with autonomous vehicles that’s a challenge- the big difference is that it required me to get a response from other road users rather than just reacting

    That’s a very good point actually – self-driving cars aside, how good would it be if a car in trouble could actively notify other cars in the immediate area? Goodbye multi-car pileups for a start.

    The-Swedish-Chef
    Free Member

    That’s a very good point actually – self-driving cars aside, how good would it be if a car in trouble could actively notify other cars in the immediate area?

    This is here now. Volvo’s can communicate skidding to other Volvo’s based on location. This is up and running in the Nordics. If you enter an area where one or more cars has reported reduced road friction then you’ll see a notification pop up on the dash.

    As always the industry standardisation of these messages and algorithms and adoption will take time.

    Northwind
    Full Member

    Cougar – Moderator

    That’s a very good point actually – self-driving cars aside, how good would it be if a car in trouble could actively notify other cars in the immediate area? Goodbye multi-car pileups for a start.

    That’s got to be the longer term goal tbh- autonomy is fine but limited, at the end of the day it’s just automating the approach we have now. If you have predictable autonomy- ie all the cars are using the same information and decision making- then you could probably have some car equivalent of Reynold’s rules and have them do some clever mass movement… But real communication and feedback, and combined decision making, is where the big wins are. It’s just, quite a big deal, and impossible til you get everything automated, and a level of standardisation that seems pretty unthinkable to an industry that can’t even work out how to make their wheels interchangable between 2 years of the same car.

    kcr
    Free Member

    That’s a very good point actually – self-driving cars aside, how good would it be if a car in trouble could actively notify other cars in the immediate area? Goodbye multi-car pileups for a start.

    You don’t have to depend on just the car in trouble to notify everyone else. All self driving cars will be continually checking their surroundings, so if anything out of the ordinary starts happening, you will quickly get warnings broadcast from multiple sources, and every vehicle will be reacting accordingly. It will be much easier to safely bring a vehicle with a blowout to a halt in that situation.

    In a situation like the recent Westminster attack, I think autonomous vehicles would react to someone driving erratically at high speed quicker than humans would. All the vehicles in the vicinity could immediately sound horns, flash lights and broadcast warnings to the emergency services.

    The-Swedish-Chef
    Free Member

    Whilst I agree with your first paragraph kcr, (with the correct industry and political willing being in place), I have a hard time seeing the second point coming true in the near term.

    dudeofdoom
    Full Member

    total engine cutout in the 3rd lane of a motorway

    hmmm don’t forget that with the smart motorway signs the motorway is also capable of interacting wiv the car in emergencies, closing lanes around it and changing speed limits etc.

    kcr
    Free Member

    Whilst I agree with your first paragraph kcr, (with the correct industry and political willing being in place), I have a hard time seeing the second point coming true in the near term.

    If you can do the difficult first bit, and build a vehicle that read the road conditions around it, and react appropriately, the second bit is trivial.

    The-Swedish-Chef
    Free Member

    the second bit is trivial.

    I don’t argue that at all, I just query if that’s what consumers and manufactures desire?

    As it happens I’m working on the use case for automatic lane changes based on blue light emergency vehicle information at the moment. So not that dissimilar to what we’re discussing.

Viewing 26 posts - 81 through 106 (of 106 total)

The topic ‘Self Driving Cars?!’ is closed to new replies.