Abstract:

Economies are facing various challenges such as increasing complexity as well as the impact of a global skills shortage. Technologies like Augmented Reality (AR) are promising technologies to address these challenges as a method of vivid, effective and illustrative (further) education and are already considered attractive solutions for on-the-job training in assembly and maintenance. However, AR is not widespread in the industrial sector. This is due, for example, to high perceived costs and low digital competencies in SMEs.

In this case study, AR instructions for tablet that are as easy to use as possible were created for the assembly of a coating unit for gas turbine blades under high staff turnover in a manufacturing SME. They were compared with paper instructions in a practical test. The goal was to holistically investigate a) which form of instructions is the more suitable for the use case and b) how the groups handle the instructions, what challenges arise in dealing with them and whether implicit, physical learning processes were observable. Due to high turnover, targeted on-the-job training is not possible. Based on the assumption that learning occurs even when it is not intended, the focus was on experiential and physical learning.

In order to determine the more suitable instruction, both were tested by one group each for efficiency (time), effectiveness (frequency & type of errors) and perceived ease of use (SUS & TAM). The results of the mixed-methods study of observation, questionnaires, as well as interviews show that the instructions differ primarily in the number and quality of errors committed. The paper group not only made more errors, but also more critical ones that would have affected the startup of the coating unit. The greatest challenge for the paper group was its own planning and independent redesign of the assembly process, which led to various critical errors.

The AR group adhered strongly to the linear flow of instructions. Their biggest challenge was handling the AR application. It was not intuitively easy enough to use, so the group relied mainly on embedded images and videos and did not use the AR-specific features.
Both groups showed signs of informal, physical learning through an initial hesitant approach to unfamiliar activities that turned into visible growing confidence, with the AR group being more successful overall. One might assume that the paper instructions left more opportunity for autonomous decision-making, as the paper group chose its own sequence of steps. However, the AR group also demonstrated active thinking rather than passive execution by suggesting a more practical sequence of steps in the process.

According to the findings, a low-threshold form of digital instructions that takes advantage of AR-functions but is more intuitive to use would be more useful. Not only would this be easier for the end user to use, but it would also be easier for people without IT skills in SMEs to create. It can be assumed that digital instructions have the potential to promote faster comprehension through clear visualizations of complex processes and objects. Thus, AR and digital instructions could be more helpful in closing larger experience gaps through vivid visualization. Furthermore, it becomes clear that AR technologies face limitations if the creators and the users do not have the necessary digital skills to operate them, especially if technical briefings are not possible.

Keywords:

Augmented reality, digital learning, vocational education, on the job learning, informal learning, experience learning, digital competencies, usability.