Automated vehicles are like vaccines - when enough of the driving population is automated, all the vehicles will move predictably and it will be incredibly safe. When only a small part of the herd is immunized from human stupidity, humans will make automation’s job much much harder.
Most of the scenarios described by the anti-automation folk above start with a human doing something unpredictable. Your automated truck can’t keep a safe stopping distance with the cars in front because greedy humans will cut into its space. This is solved when the whole herd moves by the same set of rules. I have no doubt it will not only be much safer, but we’ll have much higher capacity on roads (less effective traffiic) because we’ll avoid the traffic accordion effect. The question is how we transition there, as we can’t transition from 100% human to 100% automated in one step, so we’re stuck going through the hardest scenario.
In the future we will all have flying cars and so freeways won’t be a problem. No huh. Then maybe we will have mass transportation that is will carry everyone and will be everywhere so that private cars won’t be needed, No huh. How about hi-speed rail? Yea…right. Robot cars will be running the freeways in California before hi-speed rail will get you from LA to SF.
A lot of automated car advocates skip this transition period. They only talk about the benefits of the 100% automated roadway. Computers can react faster. They don’t make as many mistakes. When all the cars are robots, they don’t cause traffic via unnecessary braking. Etc…
The issue is that there’s going to be a period of time in which some portion of cars are automated while others are not. That’s not something you can realistically skip. That means the tech has to account for that. Roads will be shared by fallible, unpredictable humans, and flawed, WIP robo-drivers.
I think this is more of a fantasy than reality. It’s true that automated driving will be much safer, because the algorithms can see more at once and won’t get tired, bored, or distracted. It’s also true that the more automated cars there are the better they can act, because they’d be able to communicate with each other to reduce traffic, react to unexpected incidents, etc. You could even envision a situation where automated cars have been proven so much safer than human-controlled ones that driver’s licenses become massively restricted with levels of qualification akin to a pilot’s license or a security clearance. This is especially true if we keep seeing trucks used directly as weapons. Even in that case, though, the cars will need to know how to deal with unpredictable movement and rapidly evolving situations, not just to be able to respond to the occasional human doing something wacky, but because the system itself is chaotic. One disturbance will cause all sorts of reactions, and unless the cars are all hooked up in some perambulator-style control system with massive processing power and high signal bandwidth, their reactions will add to the chaos and create results as unpredictable as in current accidents. Those disturbances might be rarer because the cars will be better at safe driving, but I think lots of people over-play the idea that all the randomness is caused by human agency.
[quote=“Telefrog, post:83, topic:127557”]
A lot of automated car advocates skip this transition period. They only talk about the benefits of the 100% automated roadway.[/quote]
The converse is also true - there is a contingent of folks that toss up their hands and declare an all-automated future impossible because the transition period would be so chaotic.
The reality will probably be some weird set of cobbled together compromises until the 100% transition arrives: city-centers that are off-limits to human-controlled cars; “express” lanes that are converted to be robots-only; or the situation where you control the car manually in some situations but control is taken away from you in others.
This latter case is what I’d like to see sooner rather than later: you control your vehicle on residential or business roads, but if you get on the highway the car takes over and removes control from the human, restoring it when it reaches your designated exit ramp.
Unless there is a government mandate at some point or driving normal cars becomes harder due to driving laws, I would wager we have quite a long time where the roads will be shared.
Things that would speed it up:
Laws that make things tough on normal drivers, or force compliance in some way
Sky high insurance costs for remaining normal drivers
Reduction of lanes/roads/paths for normal drivers
Parts needed for systems for a driver may become less available, though I have a hard time picturing just what those would be beyond steering wheels, instrumentation, mirrors, pedals, etc
Sure. Frankly, I’d rather we go with a more cautious “this is going to be really hard” approach over the enthusiastic and naive sales pitches I’ve seen.
Well if this gets you to something akin to Tesla autopilot mode for $999 and it were to take off, I’m going to assume that a full system akin to what Google, Uber and others are trying would be in the thousands, not tens of thousands. But that’s just counting hardware, nothing else.
You’re gonna need crazy insurance premiums or something; there’s a substantial subset of the population (young males) that love to drive. And so long as those Fast/Furious movies keep making boatloads of money, I don’t see impressionable young males giving up their “God-given right” to the wheel.
I do see the trucking industry shifting quickly though, mainly due to costs.
Think about it. Drivers can only drive so many hours per week, and they have to log those hours. That’s government regulation because exhausted truck drivers kill people. But if they can only drive so many hours, they have to drive as fast as possible. This burns fuel.
Now imagine a convey of driverless trailer rigs. They can cruise at 45 mph, pretty much the optimal speed in terms of fuel efficiency. And since they can drive 24/7, they can get to the destination faster than a human driver who goes faster but has to stop for food and sleep. So the trucking company saves on driver salary, benefits, and tons on fuel.
But there are something like 3.5 million commercial truck drivers in the US. Throw most of them out of a job. Then throw out a lot of jobs that support those drivers, such as the folks who work at truck stops and motels. Then throw out all the jobs that depend on those jobs. And a lot of these jobs are in places in the middle of nowhere, so there’s not a lot of alternatives out there.
Don’t forget the buggy whip manufacturers. Widespread adoption of automobiles will destroy the horse & buggy industry. This is surely a terrible idea and we should stop it at all costs!
To be clear, this is not me. I just think it’s crazy to start with some of the heaviest and most dangerous vehicles on the road because some individuals believe driving trucks is somehow easy. What it really is a cost savings move for large transportation companies to try and keep their vehicles on the constantly running using the cheapest labor possible… which has nothing to do with safety. In case I am not clear, I don’t think we should start with commuter buses, school buses, or dump trucks either.
We’ve finally hit the crunch point for automation, I feel. The trend that has been building for a while is going to reach a crescendo soon. As a society we need to acknowledge, then come to grips with the consequences of, how automation of this type will drastically decrease employment in anything but certain types of skilled labor.
Finding a way for society to provide for the millions of people rendered unemployable is going to be a problem we need to come to grips with. Before heads wind up on pikes.
Not really even correct honestly. How many buggy whip manufacturers were there? Like ever?
Compare that to how many truck drivers would be put out of a job.
No one here is saying we should halt all progress on automated vehicles. At the same time these are legitimate issues that shouldn’t be passed off with a shrug.
And it’s not like those drivers will be able to move on to other driving jobs. When oil lamps go out of use, the oil lamp makers can make whatever replaced them. When you replace the workers well those guys are just thrown in the garbage.
There is a big difference between a product becoming obsolete and people becoming obsolete. You can always shift production to the New Thing, but your workers don’t magically get other jobs and skill sets.
This doesn’t sound like something that increases rate of return on equity.
Surely it will be easier to renegotiate with the insurers to offer a favourable premium despite the payouts for robotruck fatalities? Payouts are cheaper than a country wide network of backup drivers.
Or better still, Trump can install his buddies in the regulators and push through laws that remove liability between robotrucks and transport companies and increase dividends across all the logistics companies he invested in.
And we really have reached the tipping point, it seems, where those concerns can no longer be pushed off. For so long the refrain was ‘the increased growth will create enough new opportunities that this won’t matter’, but as automation scales up, and new growth slows, this no longer is really going to be the case.
It’s like how people talk about coal jobs, as if there was a way to come back. Well automation means that they can produce far more coal, with an ever decreasing number of workers. There is no way to ramp back employment there, because coal had continued to grow until natural gas cut into it in recent years. But, flat out, they are getting more production with fewer workers. Even if there was a sudden doubling of coal production the number of people working in coal would still fall far far below what it was in 1980.
Genie is out of the bottle. Learning to deal with it is the trick.
I just want to say that I never said that I thought truck driving was easy. I don’t even know if you can really say they are starting with trucks since I believe autonomous semis were talked about and demonstrated way after autonomous cars were first on the roads.
Semis being the most likely to be fully automated first is just the most likely outcome for economic reasons, as it will immediately save money for companies employing them. The same can’t be said about cars.
I can’t remember where I saw it but not too long ago I saw a graph that plotted both manufacturing output and manufacturing employment levels, and the US peaked on manufacturing output in 2007 (and we are almost back at that level) but as output increased manufacturing steadily decreased.
It’s going to be an interesting few decades as I do believe this is going to get more and more compounded, and not just in traditional areas we think of like manufacturing. Amazon is already proving that at a small scale you can have cashier-less grocery stores without the PITA that is self checkout systems.
I didn’t say you said it. I certainly did say this is basically to save money, screw the bikes, motorcycles and anyone else even the tiniest mistakes will take out because of it though… seems to be the response to hey maybe we shouldn’t use the most dangerous vehicles on the road until we can be more consistent with the smaller ones.
