An evaluation of the HumanMachine Interface for anautonomous road vehicle : A simulator study with focus on cognitive ergonomics

Abstract: Summary Autonomous vehicles are a growing trend in society and are anticipated to change the wholetransportation system as we know it today. Many vehicle manufacturers are focusing ondevelopment of different kinds of self-driving vehicles. The variety of automation in vehiclescan be divided into six levels of automation, according to the Society for Automotive Engineers(SAE) International’s division. As the level of automation varies between vehicles, the level ofneeded driver action also varies. As the automation level gets higher the drivers role changesmore and more towards being a supervisory controller for the automation. This leads to adifferent need of information presented to the driver in an autonomous vehicle compared towhat is provided in a manual vehicle. The driver must be able to understand what the automationis doing and predict how it will behave in different situations. Hence, the design of the Human-Machine Interface (HMI) is utterly important for the safety of autonomous driving. The main objective for the thesis work was to evaluate the driving compartment, from acognitive ergonomics point of view, for a special type of professional autonomous road vehicle. A literature study focusing on evaluation and testing of HMIs in autonomous vehicles wereperformed and the evaluation was done through user tests in a simulator with 22 professionaldrivers. The tests were followed by questionnaires and interviews about the test participants’experience of the designed HMI. In the literature study, only a few publications, were found addressing the validation of thewhole human machine interaction in autonomous vehicles. Most of the references had focus ontesting of limited subparts of the HMI, such as trust in automation, situational awareness andtransitions between manual and automatic driving mode. The designed HMI worked as intended, since the test participants noticed and acted upon allHMI signals and were able to takeover and handover the control from/to the automation. Somerecommendations were, however, provided for the future development work of the evaluatedautonomous vehicle. For example, since there were some limitations in the simulator used andin the test design, a next iteration of the test was recommended for future development. This evaluation was done as a first exploratory test and the aim of the thesis work has beenfulfilled. The design of the HMI, the manoeuvring devices for the autonomous vehicle and theconcept of two driver roles are, at this stage, considered adequate for continuing the project. It was also concluded that the driving behaviour of the automation is a crucial aspect forfacilitation of user-trust in the autonomous vehicle and that clear procedures for communicationare needed to ensure traffic safety. Further, it was concluded that standardization of frameworks and methods for evaluatingautonomous vehicles is beneficial for developers, authorities as well as end-users.