A pandemic, geopolitical conflicts, and trade wars have disrupted well-established supply chains in the past years. Manufacturing sites are affected down to the shop floor, facing material shortages and supply delays. Thus forcing production processes and the workforce to adapt constantly. At the same time, artificial intelligence technologies matured from niche applications to mass products, providing intelligent and self-learning tools for innovative digitalization approaches. This dissertation targets the workforce's adaptation and flexibility problems with intelligent support. It investigates the opportunity of artificial intelligence-based gesture recognition to enable situation-aware assistance in assembly and maintenance operations.
The vision is to enable anyone to fulfill a given work task and provide just the right amount of information to provide just the right amount of information, as both information overload and scarcity can negatively impact efficiency. The thesis hypothesizes that an artificial intelligence approach, combined with data gloves, provides the necessary technologies to implement such a situation-aware assistance concept. Intelligent gesture recognition has been successfully demonstrated in conjunction with camera systems. However, vision systems are prone to occlusion and reflections of light. This work explores the potential of assistance through a glove-based recognition system.