Detection, prediction, and avoidance of dynamic obstacles in urban environments
read more
Citations
Edge Computing for Autonomous Driving: Opportunities and Challenges
Local Path Planning for Off-Road Autonomous Driving With Avoidance of Static Obstacles
Model-Based Threat Assessment for Avoiding Arbitrary Vehicle Collisions
Motion planning in urban environments
Detection, segmentation and classification of 3D urban objects using mathematical morphology and supervised learning
References
Motion Planning in Dynamic Environments Using Velocity Obstacles
Heuristic Motion Planning in Dynamic Environments Using Velocity Obstacles
Randomized Kinodynamic Motion Planning with Moving Obstacles
Optimal Rough Terrain Trajectory Generation for Wheeled Mobile Robots
Related Papers (5)
Autonomous driving in urban environments: Boss and the Urban Challenge
Stanley: The Robot that Won the DARPA Grand Challenge
Junior: The Stanford entry in the Urban Challenge
Frequently Asked Questions (12)
Q2. What have the authors stated for future works in "Detection, prediction, and avoidance of dynamic obstacles in urban environments" ?
Future research will investigate how this approach can be adapted to commercial driver assistance systems with a human-driven vehicle. In particular, the approach seems well suited to intersection assistance systems, where the road structure and features in the environment can be used to provide prior information for intelligent prediction. Their testing thus far has shown that the presented collision avoidance approach can be effectively used in these and other scenarios as an additional safety layer below higher-level reasoning algorithms.
Q3. What is the main component of the Boss perception system?
Boss’ perception system provides four principle pieces of information: a vectorized road structure, a static obstacle map, an instantaneous obstacle map and a dynamic obstacle list.
Q4. What is the common approach to detect dynamic obstacles?
To reliably detect dynamic obstacles the authors use a multisensor approach combining radar and laser data from different sensors and sensor technologies.
Q5. What is the importance of predicting the future motion of an obstacle?
If an obstacle has been detected as moving, it is important to predict its future motion so that actions can be selected that are safe through time.
Q6. How can the authors modify the behavior of their vehicle to generate different candidate trajectories?
If another vehicle is detected and predicted to interfere with some of their candidate trajectories, the authors can modify the behavior of their vehicle to generate different candidate trajectories that are offset to the right of the other vehicle.
Q7. What is the way to avoid dynamic obstacles?
4.Generating predictions for dynamic obstacles traveling on roads is only part of the solution, however, since urban driving also involves navigating through parking lots and open, unstructured areas.
Q8. How many kilometers of testing have the authors done?
The authors have implemented it on an autonomous passenger vehicle and have found it to be very effective over the course of1The authors assume the time-step used for stepping along the trajectories is sufficiently small (in their case the authors set it to correspond to a distance of 0.2m along the candidate trajectory)several thousand kilometers of testing.
Q9. What is the purpose of the trajectories?
These trajectories are generated using a model-based trajectory generatordeveloped by Howard and Kelly [19] that incorporates a high-fidelity vehicle model to produce an accurate prediction of the vehicle’s movement as it executes the trajectory.
Q10. What is the method of determining the distance between the two trajectories?
The authors then check to see if these bounding boxes overlap: if they don’t, then the two trajectories cannot intersect each other; if they do, then there is a chance the trajectories intersect and the authors must continue to investigate.
Q11. How can the authors use the presented collision avoidance approach?
Their testing thus far has shown that the presented collision avoidance approach can be effectively used in these and other scenarios as an additional safety layer below higher-level reasoning algorithms.
Q12. What is the way to predict the future behavior of a vehicle?
this provides a conservative prediction of the future behavior of the vehicle (obviously, it could only actually travel down one of these lanes), but because intersections are typically prone to confusion and accidents, the authors feel that exhibiting extra caution in these areas is prudent.