The development of highly dynamic bipedal robots demands sensing capable of capturing key contact-related signals in real time, particularly the Center of Pressure (CoP). CoP is fundamental for locomotion control and state estimation and is also of interest in biomedical applications such as gait analysis and lower-limb assistive devices. To enable reliable CoP acquisition under dynamic walking, this paper presents a foot-mounted measurement system and an online self-calibration method that adapts sensor scale and bias parameters during locomotion using both external foot sensors and the robot’s proprioceptive measurements. We demonstrate an online self-calibration pipeline that updates foot-sensor scale and bias parameters during a walking experiment on a NAO-V5 platform using a sliding window optimization. The reported results indicate improved within-trial consistency relative to an offline-calibrated reference baseline under the tested walking conditions. In addition, the framework reconstructs a digitized estimate of the vertical ground reaction force (vGRF) from load-cell readings; due to ADC quantization and the discrete offline calibration dataset, the vGRF signal may exhibit stepwise behavior and should be interpreted as a reconstructed (digitized) quantity rather than laboratory-grade continuous force metrology. Overall, the proposed sensing-and-calibration pipeline offers a practical solution for dynamic CoP acquisition with low-cost hardware.

In view of the problem that high-voltage transmission lines have poor real-time performance and are difficult to effectively respond to sudden wildfires (smoke) and illegal construction (engineering vehicles) offline, this paper aims to develop a high-voltage line alarm robot system to achieve automatic, accurate, and real-time recognition and alarm of smoke and engineering vehicles in its corridor. In terms of methods, the system is equipped with multiple sensors for environmental perception and innovatively adopts the improved YOLOv8 target detection algorithm as the core recognition engine: first, rich input is constructed by fusing multimodal data; second, the attention mechanism CBAM (Convolutional Block Attention Module) module is introduced to enhance the model's ability to focus on key features and significantly improve the discrimination under complex backgrounds; third, transfer learning and domain adaptation technology are used to use large-scale pre-trained models and fine-tune with small sample data of specific high-voltage corridor scenes, which greatly improves the generalization and robustness of the model; finally, through model pruning and quantization, the optimized lightweight model is deployed to the Edge TPU edge computing unit equipped with the robot to achieve local real-time reasoning. In the real high-voltage line scene test set, the system has a recognition rate of 97.1% ([email protected]) in smoke detection and thick smoke, and the PR curve of engineering vehicles maintains F1>0.92 in the threshold range of 0.45-0.65; the average false alarm rate is controlled at 1.03%, and the average response time is less than 0.9 seconds. This robot system effectively solves the problem of real-time, accurate and automated monitoring of smoke and engineering vehicles in high-voltage line corridors, and significantly improves the efficiency and reliability of line safety protection.