Interactive Augmented Reality-guided Maintenance Operation

Training employees to troubleshoot, repair and maintain trains is a highly demanding and critical process, which requires profound technical knowledge of the train’s systems and how they interact. As infrastructure equipment and maintenance processes become more sophisticated, working with paper manuals and data sheets becomes increasingly challenging. To ensure safety and punctuality of SBB trains in the future, it is crucial to investigate means to support maintenance operators in handling this complexity.

Augmented Reality (AR) offers novel possibilities for assisting operators during manual tasks by making relevant information accessible in an intuitive manner and by reducing the complexity to manageable increments. A digital support of the maintained process has the potential to improve the process on several levels in regard to efficiency, knowledge transfer, failure reduction and documentation. AR can therefore assist employees when interacting with sophisticated equipment, dealing with elaborate processes and working in challenging environments.

In the proposed project “Interactive Augmented Reality-guided Maintenance Operation (IARMO)” we aim at investigating the benefits and implementation hurdles of AR instructions during complex maintenance procedures at SBB. We will investigate preventive maintenance tasks based on digital checklists and guidance, as well as curative maintenance for unforeseen cases of damage that require efficient identification of the problem and the right measures.

In addition, we will investigate a new generation of context-aware AR support system that combines AR instructions with real-time behavior recognition to provide optimal support. Our proposed system can understand and instinctively react as the operator progresses through the maintenance process and, for example, display the next step if a step has been completed or display additional information if the user struggles to complete a step.

Our project provides recommendations on which AR hardware and level of support are most appropriate and contributes to research on context-aware AR support in real-world applications.