https://www.mdu.se/

Purpose and research questions

Based on the large range (both in number and models) of robots that ABB Robotics provides, they also offer through their service organization different types of service of robots such as maintenance and repairs. The company is at a stage where they are broadly investigating how technologies such as Augmented Reality (AR) can be implemented to improve different parts of the business where one of them is maintenance. The purpose of this thesis will therefore be to investigate whether AR-based maintenance instructions can improve the quality and efficiency of the maintenance procedures over the currently used technique SimStructions (3D simulation instructions that are available online and can be viewed on either a PC or a tablet). The following three questions will therefore need to be answered:

RQ1: Can AR-instructions streamline the maintenance work for field technicians?

RQ2: Can AR-instructions improve quality of maintenance work for field technicians?

RQ3: Can AR-based maintenance instructions be used instead of, or as a complement to, currently used method?

Methodology 

In order to be able to answer the research questions, a case study was carried out at a company where tests were carried out for a selected part of robot maintenance using two different methods: SimStructions and AR. Furthermore, quantitative as well as qualitative data have been collected through semi-structured and unstructured interviews, observations, measurements and product manuals in the form of PDF containing detailed maintenance instructions for robot IRB 6700. To link the case study to the theoretical reference framework, a literature study has been conducted, where information has been collected from both books and scientific articles.

Results

Results from the literature study and case study show that AR can have a positive impact both in terms of efficiency and quality, as well as on learning. Tests show that the time for completing tasks was on average quite similar to a few minutes advantage for SimStructions, but where AR had the best time and that no mistakes were made with AR while there were two with SimStructions. Based on questions asked to the participants regarding the use of the two methods, information via AR led to better focus and locating ability, while SimStructions would be advantageous if individual steps were to be checked before work. The results of the experiment thus show some advantages and disadvantages, both regarding AR technology and SimStructions.

Conclusions and Recommendations 

Conclusions that can be drawn from the results are that AR-based instructions can be used as a complement or instead of SimStructions both in terms of efficiency but also from a quality point of view. However, based on the small differences that tests have shown, this would not be economically justified. However, this is based on the scarce data that could be obtained via a limited number of tests and on a limited number of test persons thus it has been a challenge to derive a firm recommendation. In conclusion, recommendations are given to perform tests on a more complex part of the robot that takes less time to perform, which enables more tests. Further recommendations consist of investigating the possibilities with training via AR, but then with a projector as several participants can see the virtual information at the same time.