Self-Driving Cars

Self-Driving Cars Finally Ready to Hit the Roads?

In recent years, many major car manufacturers have made announcements about releasing self-driving cars. Tesla CEO has claimed that they will have self-driving cars within 3 years, Nissan has announced that they will have them by 2020, and Google has claimed they will be producing self-driving cars by 2018. Clearly this idea is not as farfetched as may seem because many major car manufacturers are planning to be producing such cars in the coming years. Some manufacturers have even claimed to have tested self-driving cars successfully, a German manufacturer Daimler AG has had a self-driving car complete a 62 mile trip through the streets of Germany. Self-driving cars are not as impractical as one might think, they have been being tested for years and would likely reduce car accidents by eliminating human error altogether.
How do they work?
These types of self-driving cars are equipped with a GPS unit, an inertial navigation system, and a wide range of sensors. These sensors include: laser rangefinders, radars, and video monitors. The car uses the GPS to locate its position, the inertial navigation system to localize itself, and the sensors to monitor its surrounding environment and create a 3D image it. These cars make intelligent decisions by maintaining an internal map of its location and using that map to find a path to the destination safely while avoiding other cars or obstacles. The vehicle's map contains the locations of all driving signs such as stop signs/lights, crosswalks, and other street signs. Each car has a "Path Planning Algorithm" in which the car is only able to perform safe tasks. For example, cars will not be able to perform immediate turns while traveling 50mph, instead they will know the distance until the turn and appropriately slow down before the turn. This process of localization (understanding where the vehicle is), mapping, obstacle and threat detections, and path planning is repeated until the vehicle reaches the destination.
Application to Class: 
This is very similar to the way that we are programming the EV3 robots to do tasks we want and navigate a maze. Obviously this is on a much larger scale, but with these cars there is an algorithm and a set of steps that the car must follow to reach the destination safely. The self-driving cars are also programmed in the same way that we are programming our robots, but on a much more complex level. Just as we are programming our robot to navigate a maze, car manufacturers are programming their cars to navigate the streets in our world.