Dom, cars don't fly! -- Or do they? In-Air Vehicle Maneuver for High-Speed Off-Road Navigation

2025-03-24

AI Summary

In high-speed off-road navigation, frequent vehicle takeoffs invalidate conventional ground-contact assumptions, undermining traditional control paradigms reliant solely on error feedback. Method: This paper proposes a hybrid forward kinematic-dynamic model that synergistically integrates first-principles physics with data-driven compensation, enabling motion planning over fixed-time horizons. During brief airborne phases, the model permits active regulation of aerial attitude and landing pose—including position, orientation angles, and their derivatives—using only standard throttle and steering commands. Contribution/Results: Validated through extensive indoor and outdoor real-world takeoff experiments, the system achieves centimeter-level landing accuracy and millisecond-scale attitude response, significantly outperforming purely error-driven approaches. Its core contribution is the first integration of an interpretable physical model with learned residual compensation, establishing a deployable modeling and planning framework for “flight-style” autonomous navigation in unstructured high-speed off-road environments.

Technology Category

• Planning, Routing, and Scheduling: Model-Based Reasoning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Low Level & Physics-based Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

When pushing the speed limit for aggressive off-road navigation on uneven terrain, it is inevitable that vehicles may become airborne from time to time. During time-sensitive tasks, being able to fly over challenging terrain can also save time, instead of cautiously circumventing or slowly negotiating through. However, most off-road autonomy systems operate under the assumption that the vehicles are always on the ground and therefore limit operational speed. In this paper, we present a novel approach for in-air vehicle maneuver during high-speed off-road navigation. Based on a hybrid forward kinodynamic model using both physics principles and machine learning, our fixed-horizon, sampling-based motion planner ensures accurate vehicle landing poses and their derivatives within a short airborne time window using vehicle throttle and steering commands. We test our approach in extensive in-air experiments both indoors and outdoors, compare it against an error-driven control method, and demonstrate that precise and timely in-air vehicle maneuver is possible through existing ground vehicle controls.

Problem

• Enabling vehicles to maneuver in-air during high-speed off-road navigation
• Overcoming ground-only autonomy limits for faster off-road travel
• Ensuring precise landing poses using throttle and steering controls

Innovation

• Hybrid kinodynamic model with physics and ML
• Fixed-horizon sampling-based motion planner
• Precise landing via throttle and steering control