COVID-19 and the resulting restrictions have had a massive impact on engineering education, particularly vocational and practical aspects of training. In this study, we present a novel mixed reality (MR) tool to simulate and guide learners through a simple fault diagnosis task of a three-phase power supply. The tool was created as a web-based application that could be accessed from budget smartphones in order to cover the majority of users. Comparisons were made between novices using MR guidance and those with more experience in the task who did not have additional guidance, finding that the novices outperformed the experts across all metrics measured. This indicates that MR could be a valuable tool to supplement traditional vocational learning methods, particularly at a time when physical access to equipment and facilities is scarce. MR has applications across the engineering industry, but the target task of a three-phase power supply was chosen as it has particular relevance to the offshore wind industry, which faces a shortage of skilled engineers and technicians in the coming years.

Augmented Reality (AR) technology makes it possible to present information in the user's line of sight, right at the point of use. This brings the capability to visualise complex information for industrial maintenance applications in an effective manner, which typically rely on paper instructions and tacit knowledge developed over time. Existing research in AR instruction manuals has already shown its potential to reduce the time taken to complete assembly tasks, as well as improving accuracy [1][2][3]. In this study, the outcomes of several aspects of AR instructions are explored and their effects on the chosen Key Performance Indicators (KPIs) of task completion time, error rate, cognitive effort and usability are assessed. A standardised AR assembly task is also described for performance comparison, and a novel AR experimental tool is presented, which takes advantage of the flexibility of internet connected peripherals, to explore various different aspects of AR app design to isolate their effects. Results of the experiments are given here, providing insight into the most effective way of delivering information and promoting interaction between user and computer, in terms of user performance and acceptance.