Many mainstream cars are now available with GPS navigation systems from the factory and after market systems are available for well under $1,000. They're generally pretty reliable about finding the shortest or fastest route to a destination, occasional trips into rivers notwithstanding. Those routes aren't always the most fuel efficient however, since you can get stuck in traffic or end up driving at higher speeds that might use more fuel. Eva Ericsson at Lund Institute of Technology in Sweden is working on new algorithms to optimize routes for minimal emissions and fuel consumption. They've added fuel consumption factors for different vehicle types on various streets in a navigation system database. They also tried adding some real time traffic data but didn't have sufficient data to make it effective. Nonetheless, testing showed an improvement of 8.2 percent in fuel consumption, although they expect real world savings of about half that. Getting such a system into production would be tough because of the problem of collecting the fuel consumption factors for all the streets. However, a community-based system where drivers input information could be a viable approach and certainly wouldn't hurt.