What beacons at the horizon is complete autonomy, and it starts with automotive autonomy. The idea of self driving cars was first proposed by the American government to DARPA (The Defence Advanced Research Projects Agency) which is an agency of the United States that is responsible for the development of emerging technology for the use by the military. The idea was to build autonomous vehicles that could enter into hazardous environments, and infiltrate enemy bases without any human intervention keeping the lives of soldiers safe.
DARPA brought this to a public platform and announced the first DARPA challenge. Teams had to develop completely autonomous vehicles that would drive across the Mojave desert a distance of 142 miles and cross the finish line, to add an incentive for the participants there was also a $1 million prize up for grabs for the team that finished first.
It was an open challenge where anyone could show up, and everyone did show up! The stage was set for a battle of computational prowess, yet unfortunately no team could cross the finish line. Although a winner did not emerge, a community was set up whose one goal was to develop a truly autonomous vehicle.DARPA still wanting for greater development in this field announced within a short time that in another 18 months of the first DARPA challenge a round 2 of the challenge would be set up, doubling the stakes and the cash prize. The previous contenders all showed up for round 2 with a few new faces namely Stanford University. They brought in a new technology, Machine Learning that was relatively new at that time. They trained the vehicle over the course of a year to drive in harsh terrains, control speed, manipulate steering angles and traverse around sharp curves. Stanford University took a lead over the other contenders and crossed the finish line first. This was a milestone in the development of Autonomous Vehicles. These vehicles now had the capability to learn and adapt to the terrain, yet were still blinded. This problem was overcome with the usage of LIDAR systems. These systems allowed for the environment around the vehicle to be mapped into 3D allowing for these vehicles to sense obstacles and vary the necessary parameters to allow safe traversal around these obstacles.
Now the foundational technologies were setup that would be the underlying factor in the development of commercial autonomous vehicles which then became the primary goal. Heeding to the requests of the community DARPA setup the DARPA Urban Challenge and unlike the dessert a new set of obstacles lay in front in the development of these vehicles so that they may be safe enough to drive on the public roads. The cars now had to navigate intersections, maintain speed limits and follow all the California State traffic rules as the competition was held in an abandoned airport in California. Six teams emerged as the winners in this challenge.
This gave rise to a new set of hopefuls that would advance the development of Commercial Autonomous Vehicles leading to where we are now. Big players like Google, Uber and so on have invested in the development of such vehicle showcasing that this technology will eventually at some point revolutionize the way we drive or ….the way we don’t.
However, it is not only the advancement in technology that needs to develop it is also the confidence of the general public to allow machines to drive them around. With cases popping up like Uber where during one of it’s test drive a women pedestrian was killed due to an autonomous vehicle that could not sense her presence in front of the vehicle. This leads to many unanswered questions arising over the safety of such autonomous vehicles. But in correlation human errors has a greater changes of causing an accident rather than a well trained AI machine. The odds will have to be weighed against one another but it suffices to say driverless is the future.
Second installment coming soon to this feature post with more on the tech behind these cars!