[Disclaimer: I have done no research on these topics, so if what I'm saying has already been done or is proven to be impossible, oops.]
Automated Driving
We already have cars that can park themselves. We have airplanes that fly themselves. We have vacuums that vacuum themselves. What is holding us back from having cars that park themselves? In my opinion, the widespread use of automated driving is almost inevitable.
Motivations and Benefits:
Safety
The main cause of car crashes is the driver (as the human is the weakest link in most systems built today). There will come a point when a computer driver will be more safe than the average human driver. The computer driver would always stay focused at the wheel, follow traffic laws, never fall asleep , and stay sober. If designed and implemented efficiently and effectively, an automated driver would have a significantly faster reaction than any human being and would handle unexpected obstacles and hazards faster than any human.
Once the use of automated automobiles is widespread, the cars would be able to avoid most traffic accidents by communicating with other cars over some network. For example, if a car wants to turn left on some hill where it can't see the oncoming cars, it can broadcast to cars in the immediate area that it desires to turn left. If there is a car that is speeding along, coming from the other side, that car would be able to receive the message and respond a split second later, "Don't turn yet! I'm coming!". There are a myriad of other scenarios where this type of communications would be beneficial, such as car breakdowns, yellow lights, turn on reds, backing out of driveways, etc.
Awesomeness
Just think about it. You will never need to drive that 11-hour trip to Michigan from New Jersey to visit Caleb. You could just sleep or eat or watch a movie or play video games. You will never need to worry about getting a speeding ticket again. You will never need to find directions anywhere, and you'll never get lost. You will never need to look for a parking spot again. You may never even have to find your car in the parking lot, because, with the push of a button, it will drive to you. You won't need anyone to pick you up from the airport because your car will do it itself. You won't need to have a designated drive r for parties, because your car will be doing the driving. Just think about it. The possibilities are endless.
Traffic / Time on the road
Automated cars can help avoid this [must-watch video if you haven't learned about this stuff before].
Obstacles
Public Acceptance
The trust of the public will need to be earned, and this could take as long as a decade. It won't be easy for humans to hand over their lives to a computer, and the passengers of the car will probably spend more time nervously watching the road instead of sitting back and relaxing. Strategic marketing and emphasis of government certification will be key towards surmounting the inevitable public skepticism. Needless to say, the first priority of car manufacturers will be safety, as any sort of accident on the road will be a huge setback for the acceptance of automated driving.
Car companies will need to figure out some way to make the transition smoother for people who find it difficult to hand over control of driving. I'm not sure what this will look like, but perhaps this will include some sort of intermediate step between total automation and human driving.
Security
Software and hardware security will be of the utmost importance, as any successful attack on an automobile could easily result in the death of the passengers. In addition, if a compromised automobile sends false status updates into the communication network, it could seriously stall traffic by causing other cars to make wrong or even dangerous decisions. Even if the car's driving system can be hijacked (as it will probably be), there must be safeguards in place to identify and isolate these rogue agents. This will probably be the most technically difficult of this whole automated car business, as it is effectively impossible to implement the many different components needed to autonomously drive a car without any bugs or security vulnerabilities.
Jerkiness
Most driving robots now drive really herky-jerky, and it's important that an automated driver would be able to be a smooth driver, even smoother than human drivers. This really isn't as easy as it sounds, and it could give developers quite a headache.
Predictions:
Likelihood of automated driving being widely used sometime in the future - 90%
Estimated time until widespread use - 50 years (prototype in 15 years, government-certified 10 years after that, prices drop enough for consumers after another 10 years, 5 years before it's common to see them on the road).
Smart Street Signs
The street signs in Taiwan can be pretty horrible. They can have multiple road names on the same sign, and signs are often horribly placed. It can often be pretty hard, even in the US, to see which road we are supposed to take, even if our directions are correct.
Overview
Street signs, in the future, will be like the electronic billboards that are becoming more and more common. The will electronically display their street names, along with any other necessary information, like if there is a traffic jam there, or if the road is a dead end, etc.
Giving Directions
My vision for street signs is that they would partner with the automobile's GPS and would alert the driver when it is the correct road for the driver to take. [My assumption is that in the near future, you'll just sit in your car, say where you're going, and the correct location will automatically appear on your GPS. Speech recognition is getting better and better, and it won't be long before that happens.] The street signs would probably flash or give some sort of visual signal for the driver, but then there's the problem of having many cars pass by the same intersection in a small amount of time. How would you know if the turn signal on the street sign is meant for you and not the car behind you? Every car could have some sort of a symbol and color chosen by the driver, and though a similar symbol and color may be chosen by someone else, the probability of error may be low enough to maintain a relatively high accuracy.
Just imagine. You are driving to a restaurant you've never been to in a town you are unfamiliar with. You get into your car, say "Panchero's in Ann Arbor", and head off. Your GPS takes you down some familiar roads, but then you start reaching unfamiliar territory. Your GPS tells you that you need to turn right in 500 feet, but there are a few roads, one right after another. One of the street signs begins to blink purple with a black heart on it - your symbol. You turn on that road. You approach the highway that you need to get on. You aren't sure if you're supposed to go north or south. A portion of it displays your purple background with the black heart. You get on the highway. You drive a little bit, and then see a road sign that has been slightly changed to say "Construction Ahead" below the usual "Exit 23 - Browntown". Your GPS reroutes you to the nearest exit, which displays your emblem. You get off there, and continue your journey to the legendary restaurant Panchero's that you've heard so much about.
Difficulties
Street signs are becoming increasingly irrelevant because of GPS turn-by-turn directions, would be made nearly obsolete with automated driving. Having directions and choices projected onto windshields would also make these smart street signs useless. The cost of erecting these smart street signs and then the energy consumption needed for maintenance would probably be too much for the small return.
I actually don't think this will happen, but it's just a little idea that I had.
Predictions:
Likelihood of smart street signs being widely used sometime in the future - 10% (It will just be too costly for
Estimated time until widespread use - 15 years (if it happens at all)
Another of my big ideas that apparently is already invented:
No comments:
Post a Comment