2025 Volume 38 Issue 6 Pages 119-128
Predicting the condition of machinery and performing maintenance activities in advance is essential for efficient and safe work at construction sites. If the load applied to a machine in operation is known, it is possible to predict failures based on the load conditions. However, the load on construction machinery such as hydraulic excavators changes depending on the work environment and operation, so it is necessary to construct a load estimation model that adapts to these factors. This paper proposes a load estimation modeling method that combines machine learning to determine excavator movements and database-driven modeling. Experiments using a radio-controlled excavator show that the proposed method is effective.