preview

Obstacle Avoidance Process Paper

Decent Essays

Figure 5: Adaptive vision recognizes similar-looking terrain in the distance [5]

Obstacle Avoidance Process

The obstacle avoidance process is what allows the self-driving car to identify and maneuver around objects in the environment. This automated system allows the car to react to a changing environment. The process consists of obstacle tracking and path planning.

Obstacle Tracking

To avoid static obstacles, such as curbs and parked vehicles in an urban environment, the self-driving car generates local occupancy grid maps that maintain the location of static obstacles similar to those generated in the the virtual mapping process.

The tracking of moving objects such as cars is equally important and plays a major role in urban driving. …show more content…

The radars provide a fine-tuning for close range moving object detection, further increasing the vehicle’s reliability. [6].

Path Planning

The self-driving car decides where to drive based on path planning. Path planning produces multiple trajectories to determine the one that maximizes the set of criteria. These criteria include those that minimize the risk of collision as well as those that favor the center of roads over the periphery. The car searches for paths based on two dimensions: the amount which the car adjusts its trajectory laterally and the speed at which this adjustment is carried out.

Figure 7: (a) Potential paths are created to avoid collisions with obstacles. (b) Individual choices such as lane changes are considered. (c) A complex set of potential paths in a multi-intersection situation. [7]

The car must also select the shortest path to a destination. A dynamic-programming-based global shortest path planner similar to Google Maps calculates the expected drive time to a goal location from any point in the environment. Dynamic-programming-based means that factors such as hill climbing are considered when calculating expected travel

Get Access