Evaluation of a master thesis: Creating Appropriate Trust for Autonomous Vehicles

| Personal evaluation of a Master Thesis submitted to a foreign University, that is dealing with topics in our own research field 

I chose the master thesis by Fredrick Ekman and Mikael Johansson with the title: “Creating Appropriate Trust for Autonomous Vehicles – A framework for HMI design”, as the topic has many overlaps to my research field and intended thesis topic. 

The level of design and first feeling of this thesis is positive. The cover looks attractive, the length of 76 pages (101 in total) and a structure with 9 main chapters gives a solid first impression. A basic point that I can only criticize is the formatting of the text in two columns, which made the reading and navigation through the content difficult.  

The chapters separate the phases of the work in the overview of backgrounds, literature and methodology and practical user testing. Afterwards the framework, the concept development and the validation are documented. It is completed by an example concept and the discussion of the results, including future perspectives. The outline and structure are done well, but still it is not very easy to find specific parts of the methods, unless you studied the structure thoroughly beforehand. 

The introductory part gives a fair overview of the autonomous driving context and the aims and limitations of the work. The thesis states to be limited to the scope of Level 3 automation, which is understandable, as further derivations could be made for Level 4 scenarios. 

I would assess the scope of the work as sufficient. The literature analysis excessively describes the topic of trust, showing a solid understanding of the main and most important aspects of human trust in automation and in autonomous driving. Afterwards it discusses the relevant aspects to build the framework and validates main points. 

Regarding independence and innovation, the authors worked with scientific methods throughout the whole research process of the literature and user testing.  

The user testing was carried out in an interesting way lacking a fully automated vehicle, with a simulation setup using a right-hand drive vehicle to give the left-hand test driver the autonomous ride feeling.  

A novel graph called Optimal Life Cycle of Trust was also created to visualize the change of trust through the user journey. The framework is based on the information gathered from the research, but it sums up the necessary areas of interest regarding trust thoroughly.  

This framework is one of the main results of the thesis and can be helpful for HMI designers. It is probably not a highly innovative solution but gives a solid basis for concept development. An iterative concept development was also done with the use of this framework, that gives as further results three example concepts of new user interfaces. 

The degree of communication is clear, the explanations and concept derivations are understandable. Once the reader has understood the whole methodology and structure of the work, the explanations are comprehensible. 

Regarding orthography and accuracy, the texts were written in an understandable yet scientific form. Abbreviations were explained beforehand, and no grammatical mistakes were found. 

The bibliography is not formatted in the most readable way and misses some source details at a few points, but in general it is thorough and excessive, including many books as well as other media, which appear to be reliable sources. 

Source

Ekman, Fredrick/Johansson, Mikael: Creating appropriate trust for autonomous vehicles. A framework for HMI design. Master Thesis at: Chalmers University of Technology Department of Product and Production Development, Division of Design & Human Factors, Gothenburg, Sweden (2015), 
https://publications.lib.chalmers.se/records/fulltext/226342/226342.pdf (last opened on 29.11.2022) 

Leave a Reply

Your email address will not be published. Required fields are marked *