A Data-driven Contact Estimation Method for Wheeled-Biped Robots

2024-10-16

arXiv.org

AI Summary

To address the challenges of lacking periodic gait priors and dedicated contact sensors in wheeled bipedal robots, this paper proposes a data-driven, real-time contact state estimation method. The core innovation is the design of the first Bayesian contact filter tailored for this platform, which jointly fuses online torque-based learning with IMU-derived inertial prediction—eliminating reliance on explicit gait models or hardware contact sensors. The method integrates physics-informed torque modeling, few-shot supervised learning, and a Bayesian filtering framework. It is rigorously validated in both simulation and on physical hardware: achieving significantly higher contact classification accuracy than baseline approaches while demonstrating markedly improved sample efficiency. This work establishes a robust, lightweight perception foundation for dynamic balance control in prior-free wheeled bipedal robots.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Bayesian Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Contact estimation is a key ability for limbed robots, where making and breaking contacts has a direct impact on state estimation and balance control. Existing approaches typically rely on gate-cycle priors or designated contact sensors. We design a contact estimator that is suitable for the emerging wheeled-biped robot types that do not have these features. To this end, we propose a Bayes filter in which update steps are learned from real-robot torque measurements while prediction steps rely on inertial measurements. We evaluate this approach in extensive real-robot and simulation experiments. Our method achieves better performance while being considerably more sample efficient than a comparable deep-learning baseline.

Problem

• Wheel-legged Biped Robot
• Contact Mode Estimation
• Balance Maintenance

Innovation

• Bayesian Filtering
• Gait Estimation
• Wheeled Bipedal Robots