Road Train Project

Project Sartre is another step toward driverless cars. This stuff can't come soon enough to help improve road safety and get bad drivers off the road.

[Source: AutoBlogGreen]

Pod Cars

As I've said before, the concept of autonomous vehicle travel makes sense. Personal transportation would be more effective if controlled by computers rather than people. There wouldn't be nearly as much (if any) congestion. It's not a new concept but now the idea is making a comeback in Ithaca, New York (and other places), and it will be interesting to see if the political and cultural barriers can be overcome. Technically it seems quite straightforward:

http://news.yahoo.com/s/ap/20081013/ap_on_bi_ge/podcar_city

OnStar to Track Mileage

It's a first step toward the ultimate in vehicle navigation. OnStar is reported to begin collecting information about how much vehicles equipped with the system are driven. It will be voluntary at first, with the idea that drivers could be eligible for insurance discounts. Of course, if that is possible then the opposite is also possible.

And while some people may have privacy concerns about this, cell phone location is already tracked. Plus a road network would likely need to know where the vehicles on it are at all times in order for driverless cars to operate smoothly. This is interesting technology.

Front Camera System

Another step in the direction of autonomous vehicles. Next year GM Europe will launch a front camera system that can read road signs and help detect when the vehicle is drifting from its lane. The system works at 30 frames per seconds and can read signs up to 100 meters away. That means that at 100 km/h (62.5 mph) it can "see" about 3 seconds ahead.

Acceleration Assist

As we work toward developing the driverless car, a number of advancements have to be made. Such a car would have the ability to accelerate, brake, and steer on its own. With the electrification of those systems, controlling them is not difficult. It’s the decision making process behind controlling those systems that represents an immense challenge. The complete removal of people from the driving process will have to take place in stages.

Here’s one area where automation may become adopted first. Cars with traction control have been available for years. Some models now have launch assist, designed to minimize wheelspin and maximize traction for blazing quick acceleration.

Could these systems not be used then to minimize energy use during acceleration, and maximize smoothness in heavy traffic areas? With the use of DSRC (Dedicated Short Range Communication, see Wikipedia stub here) at busy intersections, we could start to see cars networking with one another to platoon and ease congestion. If a vehicle’s intended destination is already entered in its navigation system, as it approaches a busy intersection, the network will have a good idea of where it’s headed.

Imagine you come to a stop at a red light. You set a button in your car for acceleration assist. As the light turns green, all the cars waiting for it start to move forward in synch, including yours. No delays from the accordion effect of each car waiting for the one in front to start moving before it starts moving. Each car equipped with acceleration assist would have radar or sonar to maintain a safe distance and speed from the vehicle in front, especially if the vehicle in front is not equipped with acceleration assist.

The end result is there would be less stop-and-go and traffic would flow more smoothly. The network at the intersection knows the number of vehicles there, where they’re headed, and can adjust its timing accordingly to maximize throughput. And drivers could lay off the throttle, only steering (and braking should the need arise). Think of it as cruise control for accelerating. Something like this scenario is likely to play out as we gradually shift to driverless cars.

GM to Announce Driverless Cars Within A Decade?

Time will tell if it’s premature PR-spin but the web is abuzz today with word that GM CEO Rick Wagoner is going to announce tomorrow that GM will begin to put driverless cars on the road within a decade. A bunch of blogs are saying it. Here’s the scoop at Autoblog.

As explained in a previous post, I totally believe driverless cars for transportation is inevitable. A decade time frame is much too soon though. The reason is because it’s not just a matter of building the cars that can do it.

The road infrastructure has to be there and that’s not going to change as quickly. Cars will not be able to drive themselves on every road and in all conditions safely enough for them to be deployed in large numbers until the infrastructure is “networked” and the communication between vehicles and roads is proven.

Maybe it will happen first in another country, one with fewer roads, fewer cars, and a more cohesive highway policy. Maybe it will happen in limited geographic areas (see my next post). But for these changes to take place and then be proven enough within a decade for cars to drive themselves? That would be one hell of an accomplishment – especially in light of the current state of US automobile manufacturers. I wouldn’t bet on it.

The Lexus LS460 and (Japanese-market) Toyota Prius being able to parallel park themselves was the first step. Then we’re likely going to go through phases with limited deployment in a few areas and have cars for a time that have two modes, one where they can drive themselves and another where they can still be driven, albeit with varying levels of electronic assistance for safety reasons. And those vehicles will be able to drive themselves in those deployment areas.

How to Improve Highway Safety

From a safety standpoint it is obvious that there is no way of drastically reducing the number of fatalities and injuries from highway accidents unless the drivers become much better. That simply isn’t going to happen with the human population.

There are steps that can be taken that will help, but it’s unlikely that any of these potential steps would be politically palatable. Plus they won’t make enough of a dent. The first step is to reduce the VMT (Vehicle Miles Traveled). For a given fatality or injury rate, the fewer miles driven, the fewer accidents will happen.

Discretionary travel would have to be reduced. We’d have to cut the maximum speed limit to under 55 mph (about 51% of all traffic fatalities in 2006 occurred in places with speed limits 55 mph or higher according to NHTSA). Not too many Americans would be willing to do that.

How about annual driver certification to make sure everyone who is licensed can driver properly? Taking away licenses of those who don’t pass isn’t going to look good – especially if they’re AARP members. Stiffer penalties for those who break traffic laws? Draconian penalties for DUI convictions? The end result is going to be a lot of people are not going to be able to get around unless public transportation is available. And that’s going to create a whole host of other problems.

Instead of reducing the number of people who have access to individual transportation, how about increasing it while simultaneously improving safety? Is it possible?

The way we’re headed, yes. It’s quite clear that if we want to drastically improve highway safety we have to take human drivers out of the equation. The advantages and reasons are numerous. The technical, social, and economic hurdles considerable but it will happen. Yes, the future belongs to driverless cars. You just get in, tell it where to go, sit back, and enjoy the ride.