Optimization of Sampling-Based Motion Planning in Dynamic Environments Using Neural Networks

Schomer, Philip; Huneberg, Mark Timon; Zöllner, J. Marius

doi:10.1109/iv47402.2020.9304644

Cited by 6 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vehicular motion constraints for motion forecasting problems have been studied through the works of [12][13][14], where a combination of machine learning-based approach is overlaid on top of an optimizer to solve for assurance. While these methods ensure the generation of safe trajectories, they do not incorporate feasibility guarantees within the learning process but rather as a post-processing step, which can result in unnecessary computational resource consumption.…”

Section: Related Workmentioning

confidence: 99%

Plausibility Verification for 3D Object Detectors Using Energy-Based Optimization

Vivekanandan

Maier

Zöllner

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Accurately forecasting the motion of traffic actors is crucial for the deployment of autonomous vehicles at a large scale. Current trajectory forecasting approaches primarily concentrate on optimizing a loss function with a specific metric, which can result in predictions that do not adhere to physical laws or violate external constraints. Our objective is to incorporate explicit knowledge priors that allow a network to forecast future trajectories in compliance with both the kinematic constraints of a vehicle and the geometry of the driving environment. To achieve this, we introduce a nonparametric pruning layer and attention layers to integrate the defined knowledge priors. Our proposed method is designed to ensure reachability guarantees for traffic actors in both complex and dynamic situations. By conditioning the network to follow physical laws, we can obtain accurate and safe predictions, essential for maintaining autonomous vehicles' safety and efficiency in real-world settings. In summary, this paper presents concepts that prevent off-road predictions for safe and reliable motion forecasting by incorporating knowledge priors into the training process.

show abstract

Section: Related Workmentioning

confidence: 99%

Plausibility Verification for 3D Object Detectors Using Energy-Based Optimization

Vivekanandan

Maier

Zöllner

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…An overview of common methods for motion and trajectory planning can be found in [36]. The application of neural networks also brought about new approaches to planning for autonomous vehicles, starting end-to-end approaches like ChauffeurNet [37] or hybrid approaches like [38].…”

Section: Planningmentioning

confidence: 99%

“…During optimization, a cost functional comprising efficiency, comfort and safety aspects is minimized. A modified version of the trajectory planner can be found in [38].…”

Section: Planningmentioning

confidence: 99%

Stepping Ahead with Electrified, Connected and Automated Shuttles in the Test Area Autonomous Driving BW

Ochs,

Grimm,

Doll

et al. 2023

Preprint

View full text Add to dashboard Cite

Automated shuttles that adhere to a fixed line, a so-called virtual rail, exhibit several drawbacks, including restricted flexibility in route selection, compromised passenger safety and comfort from abrupt braking incidents, and the need for safety operator interventions. The “EVA Shuttle” project (EVA) provided an innovative public transport service that bridges the first and last mile using electrical, connected, and automated driving shuttle buses. The leading-edge concept includes breaking free from fixed routes and fixed trajectories. Due to this, the shuttles can cope with challenging scenarios such as narrow streets with parked vehicles on their own. Our shuttles also can dynamically determine new routes for flexible on-demand services to optimize the capacity utilization of the shuttles. In this work, the underlying concepts are presented, and the system will be evaluated using an overall system-oriented fleet test under real conditions in the Test Area Autonomous Driving Baden-Württemberg.

show abstract

FRENETIX: A High-Performance and Modular Motion Planning Framework for Autonomous Driving

Trauth,

Moller,

Würsching

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Optimization of Sampling-Based Motion Planning in Dynamic Environments Using Neural Networks

Cited by 6 publications

References 12 publications

Plausibility Verification for 3D Object Detectors Using Energy-Based Optimization

Plausibility Verification for 3D Object Detectors Using Energy-Based Optimization

Stepping Ahead with Electrified, Connected and Automated Shuttles in the Test Area Autonomous Driving BW

FRENETIX: A High-Performance and Modular Motion Planning Framework for Autonomous Driving

Contact Info

Product

Resources

About