Recognising Worker Intentions by Assembly Step Prediction

This paper presents a novel approach to recognize a worker’s intentions for assistive systems in manual assembly. Building on previous research, it introduces a process description for assembling products, addressing the inflexibility and limitations of previous models. This enhancement enables the modeling of assemblies for a substantial amount of products. The proposed pipeline for intention recognition employs object detection, an assembly information model, and a Markov model. One major advantage of the new method is the ability to work with limited or no training data. The online learning capabilities of the model allow for adaptation to individual workers even with just one assembly. This flexibility makes this novel approach ideal for assistive systems. We also discuss about the model’s ability to suit any type of assembly product allowing offline training and guiding unskilled workers during their assembly processes.