Certainly. However as mortality rates fall and life expectancy goes up people tend to have fewer children. So you might stop at 2 or 3 when you know the (likely boys) aren’t going to wipe themselves out ages 18.
The biggest drawback to a Google car would be that while you are in it every word you say, every move you make, every breath you take would be beamed back to Google central and whenever you tried to look out a window or play some music you would first have to watch or listen to 30 seconds worth of targeted advertising. (Maybe that is how they would recoup there development costs?)
Typical Google. Can’t do anything innovative. It has to copy GM and OnStar.
There are many problems with Taxis:
[ul]
[li] Taxis must be hailed which means you’re not 100% sure when one will be available. Try deciding you want to hail a cab in New York. It’s a lot harder than you think. It’s one of the big reasons Uber is so popular.[/li][li] Taxis are expensive because it’s still a one passenger at a time situation. A cab usually takes a single ride from point to point.[/li][li] Taxis, despite the fact they’re hard to catch have a problem finding customers. It’s why if you live in New York cabs are hard to find in many places because the cabs rather wait at the airport or the train stations where passengers are waiting rather than cruising the streets.[/li][/ul]
Think of these driverless cars more like minibuses rather than taxis. They’ll pick up passengers and drop them off as they go from place to place. Call one, and like Uber, one will be available there in minutes.
Besides, no one says you can’t buy a car, and not keep it. However, imagine buying a self-driving car, and realizing that when you’re at work, you can rent out your car and earn a substantial amount of money if you let it run off while you sit in your cube. You could program it to return when you need it.
Or, if you’re out shopping and can’t find a parking spot, instead of spending 20 minutes looking for parking, just tell your car to find its own spot. Or, maybe earn some money picking up passengers while you shop (Be back in an hour!). And, if you finish early? Tell your car to meet you back at home.
If there is enough money to letting self driving cars run around loose, then companies will buy and deploy fleets to do this. This in fact is Uber’s ultimate goal. And, if you can get a car with in minutes, and have it drop you off where ever you want quickly and efficiently, people will start wondering if they need to buy one with all of the expense and pain involved.
This is actually already becoming an issue in the auto industry. New cars can record and store big data and there is debate about who owns it. Google may ultimately be the only one capable of processing it.
I don’t think people realize how revolutionary self-driving cars will be. Imagine a major urban area like London city or Manhattan passing laws that all vehicles within its limits must be self-driving and able to communicate with other vehicles. This could largely eliminate stop signs and traffic lights (except for those needed to guide pedestrians) and some speed limits could be drastically raised (like in the tunnels). Parking lots could be packed much closer together because self-driving cars could “wake up” and move out of the way for other cars. Car-pooling could potentially be easier as self-driving cars meet at way-points and exchange passengers.
You can phone a cab company, and they will dispatch one to your location. You could buy a cab and hire a driver to operate it while you’re at work or shopping. In the situation where you want to hail a cab on a busy street and all the cabs are at the airport, how would driverless cars be different? If it were profitable to have taxis run around loose, then cab companies would do this now.
We already have multi-passenger public transportation that picks up passengers and drops them off as they go from place to place. Busses either run fixed, pre-published routes, or you can reserve a spot on a shuttle that varies its route depending on demand (e.g., airport shuttle). Light-rail and commuter trains serve most large cities. If you want a vehicle that you can call and it will be there in minutes, we’re back to taxis. How would driverless cars be different?
The many potential advantages to driverless cars are almost exclusively dependent on a complete transformation from the existing system to one where all vehicles are self-driving. Individual driverless cars can only perform functions that are already available with personal cars, taxis, busses, trucks, etc. The problems you list would still be problems for driverless cars, and your solutions are available with existing technology.
I would think the near future automated cars would augment taxis and buses for the safety features. Eventually, automated vehicles would completely replace things like taxis, buses, and the like. This has been said many times in these threads that automated vehicles would likely mean the complete overhaul of the transportation system and how people get around, but it would take a long time and many generations. Today’s more automated cars are the first step. Worrying that we aren’t completely there yet doesn’t mean we shouldn’t start to take baby steps to make it happen.
I am a retired city bus driver as of last April. Before I left, I told my fellow drivers that within 10 years, once the automated cars hit the market, they should start to get training for another job as automated busses would soon come along unless they want a job as an onboard host/security guard which might not pay as good.
You are exactly right, but the fully automated car that Google is developing is not taking baby steps. Cruise control, ABS, collision warning radar, GPS devices that talk to each other and plan routes around traffic congestion; those are baby steps. Google is designing their system to be triathlon ready.
There has been some work done with cars that communicate with sensors embedded in the pavement. These sensors could be installed in existing HOV lanes and, combined with the cruise control, ABS, radar, and GPS, could allow the driver to temporarily relinquish control and resume command when leaving the HOV lane. This system could be implemented in a few areas initially and scaled up to eventually become a fully autonomous system. Cars communicating with each other and with local sensors would not need to recognize every conceivable kind of vehicle and road condition.
No, the reason you use multiple sensors is that on many of them, rain won’t cause an enormous data flood. Many types of sensor see right through rain, so they’d remove the flood of data problem. Note also that this is a problem that human drivers already have, and can’t solve.
I hope you realize that modern cars have lots of code in them already. My Prius has had two firmware updates already. It won’t be hard for the car to failover to manual control. I know the recent Google cars don’t have steering wheels, but I strongly suspect the ones they sell will.
Actually, Google will almost certainly license the technology. The Tesla plant is right across the Bay - the Google cars go past it every day.
I have a few quibbles with the Perfect Master’s column. First, I don’t see how autonomous cars will reduce parking tickets much. I suspect you’d tell it where to park and it would do so. It is not going to feed the meter for you. Apps to find open spots are going to handle that problem.
Second, while more efficient roads may cut down on gas usage, today’s electric cars are cutting into gas tax revenue already. Even assuming the cars won’t be electric, the problem will have to be faced by per mile charges instead of gas tax charges, or else road use fees.
Third, reduced traffic tickets might mean reduced police presence, either saving money or making more police available for responding to crime. Except when they decide they need to enforce carpool lanes or cellphone usage I don’t see many cops writing tickets today. But where and when I drive it would take a flying car to even get up to 65.
You pull in front of the store you want, and then tell your car to park itself. The car could park itself in a free parking spot some distance away from where you were dropped off. Or, your car could go around and pick up and drop off passengers and earn you money while you shop. The main purpose of parking meters is to limit how long people can park in high demand spaces which is why feeding the meter is illegal in many cities.
Or can it? If you let your car park itself, it’s possible that it could pay for parking. I doubt a hand is going to come out, but a spot could have NFC payment. Car parks in the space, and transfers money much the same way your car pays for your toll when you use things like EZPass.
I think we already have flying cars in the Northeast. I already see people flying low all up and down the roads.
Why does Massachusetts have a 55mph speed limit? Because that’s plenty fast when you go through a school zone.
Even more efficient than having your car park itself is to have your car shared with others. (reputation scores and competing private firms who are reviewed online for cleanliness and how well their maintain their vehicles could fix the problem of every shared car being a filthy robo-taxi)
In that case, when you leave to go somewhere, the nearest robocar of the ones parked in various spaces dispatches itself to pick you up. More efficient that way, because it means the spaces near the door are the ones that are constantly cycling in and out.
Umm, the kind of high frequency radar necessary for imaging objects the size of your car, from something the size of your car, is reflected by snow/rain (e.g. Ka band radar can see your car and detect it’s speed in good weather, but it’s useless in the rain). Infrared and sonar are still going to see attenuation and noise from precipitation, even if you can suss out deeper characteristics than the surface using them.
And as for real-time image processing and recognizing objects, the human brain is still pretty awesome when compared to a computer. When the signal/noise ratio goes up, the brain shines even more.
ETA, and this army of sensors is going to drive the car’s price up even more.
Is the autopilot/“automation dependence” problem relevant here? Example: malfunctioning sensor confuses onboard computer, computer turns the situation over to a human, essentially shrugging and saying, “Here, you deal with this.” Startled human has been lulled into complacency by lack of sufficient training, assurances that this sort of thing won’t happen, whatever. Result: Air France 447. Flying in three dimensions is a lot more complicated than driving in two, but then again pilots are supposed to be trained to very high standards and to continuously monitor the aircraft. What happens in a driverless car when the radar cuts out cuts out and the driver is eating a sandwich and doing the Jumble?
Waskov, the car pulls over to the side of the road and stops (or wherever it is safe to stop).
I hope it can do that with the radar (or whatever) out.
Quite frankly, this may kill adoption.
Simply put, edge cases like this mean you cannot reduce the risk of death in an autonomous car to zero. It could be 10 or 100 times lower than if you drove yourself, but not zero.
So people will feel that they are safer if they drive themselves because they fear dying when they are at the wheel less than being a passenger and waiting for the robot to crash and kill them (even though it is far less likely)
The human-robot partnership method sort of fixes this. You’re still at the wheel, nominally still in control if you want to be. Of course, partnership autonomous cars will probably be much less safe than fully autonomous models.