On May 2012, the Google driverless car became the first licensed autonomous car. Three years since they began research, Google’s investment in this project has proven fruitful as this technology has become possible. This advancement shows great development and progress to General Motors’ Futurama exhibit from the 1939 World’s Fair. Designed by Norman Bel Geddes, it depicted electric cars powered by circuits embedded in the roadway and controlled by radio [ (O'Toole, Dude, Where's My Driverless Car?, 2009) ].
A large portion of the recent progress in autonomous vehicle technology can be attributed to competitions held by Defense Advanced Research Projects Agency (DARPA). In one of these contests, the 2005 DARPA Challenge, competitors were tasked with building a driverless vehicle that could traverse through an unpaved desert terrain in which a group from Stanford won. In another contest, the 2007 DARPA Urban Challenge, teams were faced with a mock-up of an urban environment. A team from Carnegie Mellon University took home the grand prize.
These competitions demonstrate that the research done over the past decades indicate that autonomous vehicles are a doable technology. In fact, specific problems solved in the research and engineering for the DARPA challenges have turned into “driver assistance” technologies seen in cars today. These technologies include automatic parallel parking, blind-spot vehicle detection, adaptive cruise control, and emergency braking. Despite no apparent hurdles in implementing this kind of technology on cars today, there are still major problems with autonomous vehicle technology. According to most experts, the biggest problems in the driverless car technology lie in sensor perception and decision-making under uncertain conditions.
A fully functional autonomous car is essentially a robot whose radar detects objects at long range. Then, laser range finders, or lidar, filter the objects as they get closer. Afterwards, vision...
Please join StudyMode to read the full document