At a Glance
- The development of self-driving vehicles, particularly in the context of automated valet parking (AVP), has been a significant focus of research and development in the automotive industry and robotics field.
- Researchers at Mach Drive in Shanghai have introduced OCEAN (Openspace Collision-freE trAjectory plaNner for autonomous parking), which significantly enhances vehicles’ ability to safely navigate to parking spots without colliding with obstacles.
- OCEAN is an optimization-based trajectory planner accelerated by the Alternating Direction Method of Multiplier (ADMM), designed to be efficient and robust in various scenarios with few dynamic obstacles, addressing the limitations of previous approaches.
- Extensive simulations and real-world experiments in public parking areas validated the effectiveness of OCEAN, demonstrating its outperformance of existing methods for autonomous parking applications and its potential for deployment on low computing power platforms requiring real-time performance.
- Successful deployment of OCEAN by automotive companies could contribute to the widespread adoption of autonomous vehicle parking technologies, potentially revolutionizing the future of automated vehicle parking.
The development of self-driving vehicles has been a long-standing goal in both robotics research and the automotive industry. While progress has been made in testing and deploying these vehicles in specific environments, recent research has focused on a new concept called “automated valet parking” (AVP). This function would enable a car to autonomously navigate from the entrance of a parking area to an available parking spot. However, reliable implementation of this technology has proven to be a significant challenge.
A team of researchers at Mach Drive in Shanghai has made a breakthrough in this area by introducing OCEAN (Openspace Collision-freE trAjectory plaNner for autonomous parking). This planner, detailed in a paper pre-published on arXiv, significantly enhances vehicles’ ability to safely reach a parking spot without colliding with obstacles along the way.
The OCEAN planner is an optimization-based trajectory planner accelerated by the Alternating Direction Method of Multiplier (ADMM), designed to be efficient and robust in various scenarios with few dynamic obstacles. It overcomes the limitations of previous approaches by accurately predicting collisions and demonstrating improved performance in real-time tests.
The researchers conducted extensive simulations and real-world experiments in public parking areas to validate OCEAN’s effectiveness. Their results showed that OCEAN outperformed various existing methods for autonomous parking applications, demonstrating better system performance and the potential for deployment on low computing power platforms requiring real-time performance.
This innovative planner has the potential to revolutionize automated vehicle parking technologies and could be further improved and tested in additional real-world trials. Its successful deployment by automotive companies could contribute to the widespread adoption of autonomous vehicle parking technologies.
References
- Fadelli, I. & Tech Xplore. (2024, March 30). An optimization-based method to enhance autonomous parking. Tech Xplore; Tech Xplore. https://techxplore.com/news/2024-03-optimization-based-method-autonomous.html
- Wang, D., Lu, Y., Liu, W., Zuo, H., Xin, J., Long, X., & Jiang, Y. (2024). OCEAN: An Openspace Collision-free Trajectory Planner for Autonomous Parking Based on ADMM (arXiv:2403.05090). arXiv. https://doi.org/10.48550/arXiv.2403.05090
