Apple's iPad integrated with self-driving car, handles navigation and UI duties
Researchers at Oxford University have developed robotic technology allowing drivers to hand control of the wheel to an unlikely chauffeur, Apple's iPad.
"We are working on a low-cost 'auto drive' navigation system, that doesn't depend on GPS, done with discrete sensors that are getting cheaper all the time," Professor Paul Newman of Oxford University's Department of Engineering Science said (via Clean Technica), describing the project. "It's easy to imagine that this kind of technology could be in a car you could buy."
The Oxford project is aimed at achieving a middle ground between the human-piloted cars that exist today and the futuristic autonomous concepts being worked on by companies like Google. Instead of the cars driving themselves all the time, the driver handles some of the work, but the car occasionally prompts the driver when it knows a route, allowing the driver the option to let the car take over.
That prompt comes by way of Apple's iPad, which is positioned on the dashboard and flashes an "auto-drive" option when the system recognizes an area. Activating auto-drive switches control of the car over to an internal system that relies on cameras and lasers built into the body of the car, as well as an additional computer in the trunk. Newman says the system works due to technological leaps forward in laser mapping.
"Our approach is made possible because of advances in 3D laser mapping that enable an affordable car-based robotic system to rapidly build up a detailed picture of its surroundings," Newman said. "Because our cities don't change very quickly, robotic vehicles will know and look out for familiar structures as they pass by so that they can ask a human driver 'I know this route, do you want me to drive?'"
The iPad remains up front as the human's main means of interacting with the system, though a simple tap on the brakes is also capable of switching back to manual control.
The technology is nowhere near being ready for commercial production. Professor Newman says the group's long-term goal is to produce a system that will cost about ?100. Currently, the prototype navigation system costs ?5,000.
The team will continue testing its iPad-driven system at its base in Begbroke Science Park, near Oxford. The next step will see them teaching the system to understand complex traffic flows and make decisions on its own about routing.
"We are working on a low-cost 'auto drive' navigation system, that doesn't depend on GPS, done with discrete sensors that are getting cheaper all the time," Professor Paul Newman of Oxford University's Department of Engineering Science said (via Clean Technica), describing the project. "It's easy to imagine that this kind of technology could be in a car you could buy."
The Oxford project is aimed at achieving a middle ground between the human-piloted cars that exist today and the futuristic autonomous concepts being worked on by companies like Google. Instead of the cars driving themselves all the time, the driver handles some of the work, but the car occasionally prompts the driver when it knows a route, allowing the driver the option to let the car take over.
That prompt comes by way of Apple's iPad, which is positioned on the dashboard and flashes an "auto-drive" option when the system recognizes an area. Activating auto-drive switches control of the car over to an internal system that relies on cameras and lasers built into the body of the car, as well as an additional computer in the trunk. Newman says the system works due to technological leaps forward in laser mapping.
"Our approach is made possible because of advances in 3D laser mapping that enable an affordable car-based robotic system to rapidly build up a detailed picture of its surroundings," Newman said. "Because our cities don't change very quickly, robotic vehicles will know and look out for familiar structures as they pass by so that they can ask a human driver 'I know this route, do you want me to drive?'"
The iPad remains up front as the human's main means of interacting with the system, though a simple tap on the brakes is also capable of switching back to manual control.
The technology is nowhere near being ready for commercial production. Professor Newman says the group's long-term goal is to produce a system that will cost about ?100. Currently, the prototype navigation system costs ?5,000.
The team will continue testing its iPad-driven system at its base in Begbroke Science Park, near Oxford. The next step will see them teaching the system to understand complex traffic flows and make decisions on its own about routing.
Comments
These self-driving cars at least are much more realistic than flying cars that 50s-60s futurists seemed to think would be around by now. People can barely handle two dimensions in normal automobiles, I'd hate to see people trying to deal with three!
Also, the driver's seat isn't a very comfortable place to do much besides drive, so I'm not sure how valuable it would be for the car to take over from time to time. I suppose you could recline your seat and take a nap, but that brings us back to my first point.
Develop a system that only requires your attention at the beginning and end of the trip, and a steering wheel that folds up out of the way when it's not needed ... now we're getting somewhere!
"KITT, could you please not drive drunk?" "Sorry Michael... bURP..."
Quote:
Originally Posted by lilgto64
what happens after a fender bender and the laser isn't aimed correctly?
I don't recall such questions when Google spent a year driving an AI car around.
The problem is they're taking all of the fun out of driving a car. Do you think Ferrari are working on this? Naw.
Quote:
Originally Posted by TheUnfetteredMind
It shouldn't do any worse than many of the people out there driving now
These self-driving cars at least are much more realistic than flying cars that 50s-60s futurists seemed to think would be around by now. People can barely handle two dimensions in normal automobiles, I'd hate to see people trying to deal with three!
In fact, driverless cars would do better overall for a few reasons:
- Software would lead to proper spacing between cars, which will significantly reduce traffic jams, particularly on highways.
- Fender benders could be virtually eliminated.
- Parking would improve.
Quote:
Originally Posted by mdriftmeyer
I don't recall such questions when Google spent a year driving an AI car around.
Don't you mean "not driving an AI car around"?
Quote:
Originally Posted by lilgto64
what happens after a fender bender and the laser isn't aimed correctly?
They'll have to have mulitple sensors around the car. This stuff is STILL years away.
If you look at the Google car, they have that MEGA expensive and impractical spinning camera on top of the car and that DEFINITELY won't fly as a production model.
What's going to happen is that certain mfg will bring little bits of pieces of what they are working to control the car for short durations and that's what we'll see for a LONG LONG time. Driverless where we don't have to touch the car on short and long trips? NOPE. Not gonna happen for a LONG time.
I watched a report from the Auto industry and that's what the majors are planning, technology by teaspoon, not buckets.
Mercedes, Audi, Volve, etc. are implementing little things every so often like being able to park without the driver, braking without the driver, etc.
There's quite few taxi drivers here in NZ I could gladly murder.
They would benefit from this - you could enter the address so you don't have to speak to them.
Software does the job - not the non Engrish speaking moron behind the wheel.
Not all - but some.
I had one take me against the traffic down a one way street so he could pull a u turn to get to the on ramp. "Quicker this way." He says
Im going wtf !
lol
I believe you are correct for a number of reasons:
Quote:
Originally Posted by drblank
They'll have to have mulitple sensors around the car. This stuff is STILL years away.
If you look at the Google car, they have that MEGA expensive and impractical spinning camera on top of the car and that DEFINITELY won't fly as a production model.
What's going to happen is that certain mfg will bring little bits of pieces of what they are working to control the car for short durations and that's what we'll see for a LONG LONG time. Driverless where we don't have to touch the car on short and long trips? NOPE. Not gonna happen for a LONG time.
I watched a report from the Auto industry and that's what the majors are planning, technology by teaspoon, not buckets.
Mercedes, Audi, Volve, etc. are implementing little things every so often like being able to park without the driver, braking without the driver, etc.
Perhaps you're not thinking through the financial implications of a self-driving car. Self-driving cars are the end of car manufacturers. If my iPhone is driving my car, I don't care how well it accelerates. If my iPhone is driving my car, I don't care how well it takes corners, because the car can slow down to whatever speed is dictated by the terrain and conditions for cornering. If my iPhone is driving my car, then the car isn't really an extension of me; the car is an appliance. The ideal self-driving car is a Prius, but with a smaller engine and a smaller motor. As a result, no major manufacturer has any interest in pushing self-driving technology forward. The things that will push the technology forward are 1) very high mileage, 2) convenience, and 3) safety, and, eventually, 4) much higher speed.