Effective Detection and Classification of Robot Execution Failures using Extreme Gradient Boosting (XGboost) Algorithm
The results demonstrate that XGBoost outperforms other models in analyzing sensor data and identifying motion errors and improper robot performance, making it an effective tool in developing AI-driven robots.
Abstract
In the field of robotics and artificial intelligence, detecting and correcting motion errors in robots is crucial. In this study, the Extreme Gradient Boosting Algorithm (XGBoost) is used as a powerful tool for identifying and categorizing these errors. Sensor data, such as information from force and torque sensors, is utilized in the “Robot Execution Failures” dataset. The results demonstrate that XGBoost outperforms other models in analyzing sensor data and identifying motion errors and improper robot performance. This algorithm also enables measuring recovery from error states to normal conditions, enhancing accuracy in error detection compared to other learning models, making it an effective tool in developing AI-driven robots.