Between hiking, cycling, and doing research on transportation and tax-increment financing, the Antiplanner has been too busy to write a decent column today. So I’ll just link to a couple of recent articles on one of my favorite topics, driverless or autonomous cars.
First, the Kansas City Star notes that driverless cars are “just around the corner.” The article apparently also inspired an editorial cartoon on the subject.
Person catalysts of worry are identified with peas, lentils, chickpeas sample cialis and different beans. Specialized and classified nursing (nursing class is decided by the requirement of http://www.devensec.com/rules-regs/decregs104.html cheap online levitra each ward)- interpreters ready for intercourse. These two developed compounds are Tadalafil and Dapoxetine. viagra samples canada Their libido cialis for sale australia and hunger for sex know no bounds and they always crave for more and more.
Second, the New York Times Freakonomics blog has a second article about driverless cars. This is a follow-up to the one it had last July. Freakonomics has previously covered other transportation issues, including toll roads and gas taxes.
I would be interested in some discussion about the CPU power required for a driverless car system in a city. Human drivers in aggregate likely have a CPU power greater than that of any supercomputer. Does anyone think a driverless system would be more efficient or suffer fewer accidents than our current system? It’s not obvious to me. We’ve recently had automated rail accidents, and rail is infinitely simpler that surface traffic.
Of course, a driverless system that radically reduced driver options as to time of travel, destinations, speed, lane, spacing, exit and entrance might just work. But isn’t that a bus system?
Then, of course, what would be the costs of the needed control system: pavement embedded wires or radio controlled, how much band width, how many frequencies, are there enough available, computer system or systems, both onboard the cars and for central traffic control, programming, etc.
What is our current experience with traffic flow control?
rmsykes: Do a Google search of this site, and you’ll find many articles and discussions on the topic that might answer some of your questions. I know that GPS has been discussed as a tool, eliminating the need for any embedding in roads.
Further to that, the Freakonomics article is a pretty good summary of the state of driverless cars.
The Darpa Grand Challenge 2007 results were pretty remarkable: autonomous vehicles operating in a pretty challenging recreation of an urban environment. The best of them worked very well. They did so using off-the-shelf computing systems that fit inside of the car.
I think a well-developed driverless system would be FAR safer and more efficient than the current system. Check the safety records of automated train systems versus driven train systems. My home town’s Skytrain is the longest driverless system in the world, and it has a superb record. This is a bit of a chicken and egg, mind: nobody is likely to promote or favor a driverless car until they are notably safer than driven cars*.
The current expectation is that the control system would be onboard, and the positioning system would be GPS combined with dead reckoning (and, at a guess, maybe A-GPS and Skyhook positioning; those are existing technologies).
Brad Templeton has a good set of essays on the present and future of robocars. I’m convinced it’s a good idea and frankly eager to get to that future.
*Considering the number of accidents caused by driver impairment, fatigue, and outright recklessness, this bar might be much easier to leap than you think.