in

Love ItLove It

The arugments for all to support Driverless Cars.

One of the things I am often asked to do professionally has a future discussion. I want to be careful with how I say what’s. Next, yesterday I posted a brief note about driverless cars. The intent was to gauge the response to the concept. It is important to know what your audience believes and knows before diving into a new topic or one that you’ve covered lightly before. Driverless cars are such a topic.

But, that said, I did want to share some of the information available right now.

1. The vast majority of car accidents (well more than 60%) in the US are read end collisions.

2. The cost of those accidents drives up car insurance for everyone.

The initial arguments against driverless cars are interesting. The hardest one to deal with is fear. Everyone has a right to be afraid of change. But, there are six companies right now working on driverless cars in the US (and another 5 in Europe and 2 in Asia Pacific). Asia could have more, but the Chinese Automatacturors don’t always share information about future state products.

The estimate is that these companies have cars that have driven more than 10,000,000 total miles. Why does that matter? Well, it allows us to address concern 1.

•It is too hard to predict human behavior.

Honestly, if we start with that argument that is the REASON to have driverless cars, humans can receive input from their six senses. They are limited by the human head when it comes to vision (we see at best 240 degrees at any one time). A driverless car sees 360 degrees all the time. The second is if you are in the car with another person and they respond orally to stimuli, your natural reaction is to jump a little and check on that person. A computer doesn’t need to do that.

So, want to reduce unpredictable humans, deploy driverless cars.

•The tipping point for Driverless cars is 40% of the total car population.

That one is an estimate, and while the number may be wrong the concept isn’t. The reality of Driverless cars is, ultimately, safety.

When we consider that 18 and 19 years olds die in car accidents more than any other single thing, it is pretty scary overall.

•Humans make the software, and well humans are flawed

This one is interesting. There are inherent flaws in humans. But, just because we design something doesn’t mean it has the same flaws. A driverless car will be able to see 360 degrees on the road plain. It will be able to see ahead of the car, including the known car blind spots, and it will see above the car. Driverless cars can see more, see further and see more clearly than humans. Driverless cars will not be distracted by the noise in the car. Driverless cars do not suffer from road rage when someone cuts them off. The reality of human flaws may not, in the end, translate to the software that drives the car.

The time of driverless cars is nearly here. I, for one, cannot wait!

  • Did you know the vast number of car accidents are caused by human error?

    • Yes
    • No
  • Driverless cars can’t drink and drive. Right?

    • Yes
    • No
  • Driverless cars will still have a driver in the drivers seat, right?

    • Yes
    • No
  • sometimes we have to stop and consider hte big picture right?

    • Yes
    • No

What do you think?

20 points
Legend

Written by DocAndersen

I am a long time blogger and technology poster.I focus on what is possible, but I also try to see what is coming. In recent years I have been focused on sharing the memories of my family, as part of my Family History Project.

39 Comments

Leave a Reply
  1. You have many comments on this, Doc.
    I guess it shows this matters to them personally whatever their opinions.
    Most things have been expressed, but I am sure the ‘human driver’ will be phased out eventually unless all our resources are gone and we revert to horses!

  2. Q: Did you know the vast number of car accidents are caused by human error?
    Yes (10 votes) – 100%
    Q: Driverless cars can’t drink and drive. Right?
    Yes (10 votes) – 100%
    Q: Driverless cars will still have a driver in the drivers seat, right?
    Yes (9 votes) – 100%
    Q: sometimes we have to stop and consider hte big picture right?
    Yes (8 votes) – 89%
    No (1 votes) – 11%

    • cost is a very interesting discussion.

      I don’t think that you would pay as much for a driverless car in the future as you would today. But, that would be interesting.

      Today I spend between 200 and 300 a month for car insurance, that is actually the 2nd largest cost per month for 5 drivers in our house.

  3. It’s an interesting concept, and the fact some car manufacturers have put them into use. What if the electronic system malfunctions? What if you tell it to turn right but it turns left because if a glitch, an electronic error?
    Can it always avoid accidents? No. It’s safer but not very safe. In any case, it will act as a relief while driving on a long journey. A person can as well take a nap, letting the electronic driver drive.
    I wonder: what if it develops a mind of its own? Won’t you be engaging in an argument with it. For instance, you tell it to go a particular area but it responds it doesn’t like that place; it prefers another place.

    • your argument would require several advances in AI before it was real.

      The driverless car would in fact have a human back up system.

      I would actually argue the other side.

      1. If we were to reduce the number of rear-end collisions that would free up billions of dollars in the world currently spent on insurance.
      2. if we reduce high-risk driving (aggressing, driving while drinking, etc) that would reduce human fatalities.
      3. If no one is driving a car, when someone does, it can be remotely shut off reducing further the reality of the impact.

    • That is an interesting reality, I suspect that the easy answer would be what ultimately will you gain hacking a car?

      People always throw out “it can be hacked” but always seem to forget that major vulnerability to all systems is not the code.

      It is the people using the technology that are more often hacked.

      • Yes, a lot of hacking is social hacking. But anything on can be hacked (and even things off often have some residual power in them). As to what you would gain – well, imagine someone playing Carmageddon with real cars. Sometimes what is gained is simply the chance to cause terror and destruction. And driverless cars still have something human somewhere in the system, and it doesn’t have to be the occupants. Which means they can be socially hacked as well.

        • Ah, of course you are assuming that there wouldn’t be controls for that. Yes, today hackers are ahead, but Cyber folks are catching up. THe only hackers that could manage something that big, my gut, would be a nation-state. That would be an open act of war.

          • Well, anything that currently gets hacked does have controls to stop it. Often they don’t work (often because of human error admittedly). There’s also a tendency to not update things that don’t have a regular interface as often. There’s a reason many systems have an air gap as a major security factor.

        • first you are applying a very broad and very generic argument to this, that is not really proven.

          When is the last time you saw Microsoft, Amazon or Google get hacked directly? Because one of them is building driverless cars, the other two build in car systems.

          Controls are interesting. But under what model are you considering the Cyber Controls?

          The model is what is critical.

          • The problem isn’t direct hacking of those companies; the problem is hacking wireless networks the cars use. And currently security on those is not as great as it really needs to be. Here’s an interesting article: https://www.theguardian.com/technology/2015/mar/09/driverless-cars-safe-hackers-google

            I do believe that driverless cars, if not tampered with, are safer than human drives. They don’t get bored, distracted, fall asleep at the wheel or drive whilst drunk. I don’t believe it is possible to make them truly secure from external influence. If nothing else, somewhere along the line, a human will fail.

        • The assumption of course that they are on a wireless network that can be hacked. interesting.

          The Guardian article is reputable, but it assumes a couple of things that may or may not be true long term.

          One of the implementations that I’ve seen for driverless is to utilize blockchain within the system. Now, the hacker has to hack 4, 5 or more times. Less likely to happen (they will be found).

          • There are things they can do to improve driverless cars. Currently, I’m assuming they are on a wireless network because how else will they receive map and traffic updates? That, I think, is the biggest point of vulnerability; getting the needed information without compromising security.

            I imagine we are still years – decades? – away from driverless cars being truly accepted on roads, much less commonplace. Hopefully the problems will be sorted by then, but I cannot imagine anything we make being truly free of problems, going by all tech, ever.

            Anything that makes hacking more difficult needs to be included in every make, otherwise hackers will gravitate to those that are easier to hack.

          • It will vary from country to country, but I’m thinking more about how long it will be before it is legally approved for driverless cars to be in common use rather than for them to be technically capable.

          • I’m not sure where it stands as yet in the UK. I know that a number of organisations thought a proposal of having ten-lorry convoys on the motorway was a bad idea. There’s also a bit of difference between approving a prototype and test and approving mass usage.

Leave a Reply