Ford wants your car to avoid crashes on its own
Is driving safer if your car and that jerk's car agree not to trade paint?
There are a lot of ideas out there for cutting down the number of accidents on highways, while making traffic a lot more efficient. Vehicles remote-controlled by intelligent road networks, artificial intelligence built into the cars so they can drive autonomously, systems that honk or shock or jolt drivers to wake them up when they start to nod off.
Most require lots of infrastructure and compute power, neither of which is likely to come with one car with a reasonable price, a couple of kids and a bunch of suitcases and beach toys for vacation. Most auto-autos are so packed with sensors and gear the developers can't ride in them, and are probably healthier for it.
Lacking all that, but eager to add preventative measures that sell for good prices to safety conscious consumers, carmakers are adding intelligence to the cars themselves to avoid crashes by doing little more than notice when another car is too close and warn it to watch its step.
They don't always work. In this 2011 test a Volvo crashed during a demonstration of its crash-avoidance system.
Vehicle-to-vehicle communications systems could eliminate or reduce the damage from 4 crashes out of five that don't involve an impaired driver, according to a report from the National Highway Traffic Safety Administration.
Ford, GM, Toyota, Daimler and others have been working much of this decade on warning systems as part of various industry groups.
No single design has emerged, but Ford is putting together a set of demo models it will take on tour in the Spring and Summer to demonstrate the most-agreed-upon approach: a system with both GPS and WiFi that can identify where it is and warn cars around it to help each of them avoid the others.
Some luxury cars also have visual or radar sensors for collision avoidance or parking assistance. This approach is designed to be much less expensive and widely adopted, to let the cars themselves avoid collisions by talking to each other -- keeping a driver from sideswiping someone in the vehicle's blind spot, or hit the brakes more quickly than a human driver notices cars in front are slowing rapidly to avoid a crash or other obstacle.
Peer-to-peer avoidance systems have a lot of advantages over more sophisticated networks, and could work better because of them.
They're a lot cheaper, so it's more realistic to expect they could be deployed. Their information is always up to date because they're constantly monitoring a tiny zone around them, rather than trying to monitor and control a meta traffic pattern.
And both of the main technologies are already being built into cars.
GPS is becoming a standard feature for luxury cars and a common option on lower-priced models.
And a disturbing number of vendors are looking for ways to add real-time Internet access to moving cars, using WiFi, cellular networks or other means.
You'd think that would guarantee more mayhem from distracted drivers.
But tied with simple, peer-to-peer avoidance systems, they could end up making things a lot safer.
At least for drivers willing to not address their email until they stop the car.