[Paper Review] A Geometric Approach to On-road Motion Planning for Long and Multi-Body Heavy-Duty Vehicles

Rui Oliveira, Oskar Ljungqvist+3 more|arXiv (Cornell University)|Oct 14, 2020

Robotic Path Planning Algorithms20 references4 citations

TL;DR

This paper proposes a geometric framework to optimize motion planning for long and multi-body heavy-duty vehicles by deriving optimal trade-offs between centering different vehicle axles—front and rear axles for buses, and tractor and trailer rear axles for articulated vehicles. Using analytical geometric derivations, the method computes online weighting parameters that minimize the swept area, resulting in paths that keep the entire vehicle body centrally aligned within the lane, significantly improving safety and maneuverability over prior methods.

ABSTRACT

Driving heavy-duty vehicles, such as buses and tractor-trailer vehicles, is a difficult task in comparison to passenger cars. Most research on motion planning for autonomous vehicles has focused on passenger vehicles, and many unique challenges associated with heavy-duty vehicles remain open. However, recent works have started to tackle the particular difficulties related to on-road motion planning for buses and tractor-trailer vehicles using numerical optimization approaches. In this work, we propose a framework to design an optimization objective to be used in motion planners. Based on geometric derivations, the method finds the optimal trade-off between the conflicting objectives of centering different axles of the vehicle in the lane. For the buses, we consider the front and rear axles trade-off, whereas for articulated vehicles, we consider the tractor and trailer rear axles trade-off. Our results show that the proposed design strategy results in planned paths that considerably improve the behavior of heavy-duty vehicles by keeping the whole vehicle body in the center of the lane.

Motivation & Objective

• To address the lack of effective motion planning strategies for long and multi-body heavy-duty vehicles, such as buses and tractor-trailers, which face unique challenges not present in passenger cars.
• To resolve the conflicting objectives of centering multiple vehicle components (e.g., front and rear axles of a bus, or tractor and trailer of a semi-truck) within a single lane.
• To develop a computationally efficient, online method for determining optimal weighting parameters in motion planner objectives, avoiding time-consuming offline tuning.
• To unify motion planning for both long vehicles and multi-body vehicles under a single geometric optimization framework.
• To improve path quality by minimizing the area swept by the vehicle body, enhancing safety and stability on urban roads.

Proposed method

• Derives optimal weighting parameters for motion planner objectives using geometric analysis of vehicle kinematics and road constraints.
• Models the vehicle as a series of connected rigid bodies with defined geometric envelopes to compute the maximum lateral deviation from lane center.
• Uses stationarity principles and envelope width minimization to determine the optimal trade-off between axle centering objectives.
• Applies the derived weights in a numerical optimization framework using Sequential Quadratic Programming (SQP) and Real-Time Iteration (RTI) schemes for real-time feasibility.
• Computes the optimal weighting parameters online, enabling dynamic adaptation to changing road curvature and configuration.
• Validates the framework using real-world road data from Scania’s test facilities, ensuring practical relevance.

Experimental results

Research questions

• RQ1 How can the conflicting objectives of centering multiple axles (e.g., front and rear axles of a bus) be optimally balanced in motion planning?
• RQ2 What is the geometrically optimal trade-off between centering the tractor and trailer rear axles in articulated vehicles?
• RQ3 Can the optimal weighting parameter for motion planner objectives be derived analytically rather than through time-consuming offline tuning?
• RQ4 How does the proposed geometric optimization improve the vehicle’s swept area and lane centering compared to existing methods?
• RQ5 Can the derived optimization framework be applied online and adapted dynamically to varying road conditions?

Key findings

• The proposed method reduces the maximum lateral deviation of the vehicle envelope to 2.30 m on the left and 2.26 m on the right, closely matching the geometrically expected width of 2.27 m.
• The motion planner achieves a balanced centering of both tractor and trailer, with both envelope widths within 0.03 m of the theoretical optimal width.
• The planned path curvature closely follows the optimal curvature derived from stationarity principles, with a maximum deviation of less than 0.01 rad/m.
• The articulation angle β of the tractor-trailer is well-maintained, with the planned path closely matching the geometrically derived β profile.
• The Real-Time Iteration (RTI) scheme achieves average computational times of 0.030 s for buses and 0.050 s for tractor-trailers, with worst-case times an order of magnitude faster than SQP.
• The method eliminates the need for offline tuning of weighting parameters, enabling dynamic, online adaptation to road characteristics and vehicle configurations.