Enhancing Inclusive Manufacturing through Human-Machine Reciprocal Learning: A Trust-based Framework

This paper introduces a trust-based human-machine reciprocal learning (HMRL) framework to foster inclusive manufacturing by providing integrated cognitive support for neurodiverse workforces in human-robot collaborative (HRC) environments. The framework's architecture comprises five key modules: cognitive assessment, real-time object detection, natural language processing, a multi-modal instruction interface, and collaborative robotics. A reciprocal learning mechanism facilitates continuous improvement in both the machine learning models and the personalized assistance provided to workers with cognitive differences. Evaluation across diverse manufacturing scenarios demonstrates significant gains in assembly performance, including improved quality and reduced cycle time. This HMRL framework lays the groundwork for symbiotic human-robot manufacturing.