Inspection-aware industrial augmented reality: integrated concepts and preliminary validation for extending an AR-AI station for component recovery from washing machines

This study presents an extension of an industrial augmented reality (AR) and artificial intelligence (AI) disassembly station with inspection functions for reusable components recovered from end-of-life washing machines. While prior work focused on monitor-based AR guidance and AI-supported validation of disassembly steps, this paper addresses the next integration step: inspection-aware AR workflows that connect worker guidance, AI/sensor-based condition assessment, human confirmation, and reuse documentation. Specifically, we outline how inspection procedures for the two most valuable components targeted for reuse (printed circuit boards (PCBs) and the washing unit) can be integrated into the station concept. For PCBs, a burn-damage classification pipeline was developed. Preliminary frame-level validation on an independent real-world dataset showed that several models detect most visibly burnt PCB views, but their low specificity currently prevents autonomous reuse decisions. The AI component is therefore used as a preliminary screening and warning mechanism whose output is reviewed by the worker. For the washing unit, the developed testing device demonstrates the feasibility of pressure-loss and vibration measurements directly after disassembly; these measurements are presently proof-of-concept demonstrations rather than quantitatively validated diagnostics. The primary contribution of this Brief Research Report is therefore the system-level integration of AR guidance, visual and sensor-based assessment, human confirmation, reuse classification, and documentation into one continuous worker-facing workflow. The results establish integrated feasibility and preliminary validation rather than industrial inspection reliability.