Walk ahead of the University of Michigan, the University of Michigan, grabbed Google and other large companies before the first launch of the application of the unmanned vehicles, that is, today's protagonist MCity bus. MCity bus officially put into Sensor use last week, harvested a lot of media exposure; but as early as December 9 last year, the University of Michigan in the official website has been reported on it. In the V2V model, many unmanned cars can be connected to each other as a 'car network', the application of special short-wave communication system to ten times the speed of information sharing, such as location, speed, direction and other data.
Through the car V2V exchange, you can see other physical sensors can not see the scene, which not only can improve the safety of the vehicle being tested, but also for the surrounding other vehicles to provide a better driving environment. Another new initiative of the Mcity project team is the introduction of AR technology into critical testing sessions for unmanned vehicle iterative upgrades. According to the director of the MCity project and the professor of the University of Michigan, Professor Huei Peng, through the AR technology, the MCity project team Suction Control Valve was able to use 32-acre test sites filled with virtual cars to simulate the complex traffic environment of the city and suburbs. It is worth mentioning that these virtual cars also have the ability to communicate with the physical car. There are two advantages to using AR to test, one is to save costs to avoid the problem of car scrapped, the second is to avoid a security incident.
In the test, Lincoln test car in the red light before the accurate stop. This judgment does not come from a physical sensor such as a camera, but comes from a vehicle-infrastructure communication. Before a green light, Lincoln made his own judgment, decelerating the brakes. What is this? The original cross-direction of a car did not slow down, broke the red light; in order to avoid this sudden incident, no car using V2V quickly made the right choice. But for the use of traditional sensors for unmanned vehicles, whether in such a short time Fuel Rail Pressure Sensor to make judgments need to mark a big question mark. Found that the red light target - to carry out the perception - transmission signal - the control system to receive and determine the signal - issued a command - slow brake, this long process is likely to lead to unmanned vehicles to escape, lead to tragedy.
The biggest bottleneck in restricting the development of unmanned vehicles is the bad performance of coping with complex emergencies, and no matter how hard it is for scientists to enter every possible accident, they can not exhaust all possible. But Mcity's V2V chose another way, proposing a very good solution compared to it. Researchers only need to enter several typical events, the rest to the car through the exchange to make their own judgments can be. Despite the need for a lot of testing, countless dangerous scene simulation, the University ABS Sensor of Michigan developed V2V has taken a big step in the development of unmanned vehicles. The interconnection and autonomy between cars is the positive solution to maximize the security potential.