A team led by Prof. Paul Newman at Oxford University, UK has demonstrated their Self Driving RobotCar on the private roads of Begbroke Science Park. It proves to be a great achievement as it can take over the driving once you accept the offer from the car’s computer. Getting back the driving control is also made simple by just tapping the brakes. It is not the end, there are lots more features to know about this exciting RobotCar.
Oxford RobotCar is a modified Nissan Leaf, which has completely different features from other self-driving cars. Instead of GPS or LIDAR sensor, this car uses two stereo cameras and two scanning lasers for scene recognition. There are three computers, which act as the heart of this self-driving system.
- The first one is an iPad, which offers a self – driving mode on the known routes, and informs the details of human intervention or similar obstacles to the driver.
- A Low Level Controller (LLC) provides continuous attention to the iPad.
- A Main Vehicle Computer (MVC) does the impact work of the system.
If any one of the computers disagree the situation, then the car will request the driver to take back the control. If the driver fails to do so, then it will automatically slows and stop.
The probability and machine learning in the system helps to monitor the cars, pedestrians and other obstacles in front of it. It is capable of scanning up to a distance of 50 meters for an angle of 85 degrees. This action is performed 13 times in a second to get rid of obstacles ahead it.
This autonomous navigation system developed by Mobile Robotics Group (MRG) of Oxford University costs only US $7700, and the team believes that it can be brought down for just US $150.
The below video brings you the demonstration of Oxford’s Self-driving RobotCar: