Introduction to the Problem
Augmented reality (AR) has the potential to revolutionize industrial operations by providing real-time, interactive information overlaid onto the physical world. By integrating digital content with the physical environment, AR can guide workers through complex tasks, provide step-by-step instructions, and highlight potential hazards. For example, AR can display assembly instructions directly on or near the workpiece, reducing the need for paper manuals and minimizing the risk of human error. Additionally, AR can be employed for remote assistance, allowing experts to guide workers on-site from a distance. However, the successful deployment of AR-based systems in industrial on-the-job training and guidance is contingent upon various factors such as the availability of necessary expertise, existing digital content, and other deployment-related criteria like task complexity and error susceptibility. In particular, the suitability and appeal of AR as a worker assistance system in highly customised manufacturing environments with diverse, complex, and non-complex products and their variants often remain uncertain for decision-makers.
Solution
Figure reproduces from [1] under CC-BY 4.0
To bridge this knowledge gap, we developed a decision support tool (see Figure) tailored to craft customized deployment strategies for AR-based assistance systems, primarily focusing on manual assembly tasks. The tool was developed based on a rigorous process involving exhaustive systematic analysis of existing literature and the expertise of sixteen domain experts. Thanks to this, we were able to identify factors such as asset and task complexity and the effort required to acquire data for assistance systems to be crucial when deciding on AR deployment in the first place.
All this captured expert knowledge was utilised to develop a decision support tool to help decision-makers when considering AR deployment in their industrial use case scenarios. In consequence, we have developed a user-friendly expert system that provides a practical framework for designing and implementing effective AR-based assistance solutions. Our tools support the consideration related to the customisation of AR interfaces and assistance options to address the unique complexities of individual assets and tasks. By drawing upon a thorough review of existing literature and incorporating valuable insights from domain experts, we have created a tool that facilitates tailoring AR interfaces to support the decision-making process. This empowers decision-makers to optimise the worker experience and maximise the performance of AR systems. Additionally, our solution serves as a valuable checklist for AR system deployment, offering guidance on the available criteria and modalities to consider when developing such immersive interfaces.
More open access details can be found under this link:
https://link.springer.com/article/10.1007/s11042-024-19861-x
The Decision Support Tool is freely available at:
https://github.com/skt40/DecisionSupportForAR
[1] Bock, L., Bohné, T. & Tadeja, S.K. Decision support for augmented reality-based assistance systems deployment in industrial settings. Multimed Tools Appl (2024).
https://doi.org/10.1007/s11042-024-19861-x