Proposal for semi-automatic control reflecting operator heuristics using fuzzy logic and image detection with monocular camera images

Automation of construction machinery is being advanced to enhance the efficiency of construction work and address the shortage of skilled operators. A key challenge in automating construction machinery lies in responding to various changing external environments. To address this challenge, extant studies have sought to incorporate operator models that simulate operator actions into automatic control systems. This paper proposes a semi-automatic control system that integrates human heuristics into the boom operation of the “sweeping operation”. In the proposed method, the automatic control of the boom detects the position of the bucket that the operator is focusing on by using monocular camera images captured from the operator's perspective. Based on the operator's heuristics revealed through interviews, the operational amount of the boom is determined using the bucket's position information combined with the arm speed information. Validation results indicate that the control system successfully integrates the operator's heuristics for boom operation by considering both the position information of the bucket and the speed of the arm. Thus, a framework for semi-automatic control that effectively incorporates operator heuristics is established.