Do you remember the autonomous vehicles that travelled around the city of Sion between 2016 and 2018? These driverless shuttles have been tested around the world, and the results in terms of technology have been extensively studied. As for the human factor, i.e., operators and technicians, it has been poorly documented. To remedy this, Jakub Mlynar, a research associate at the Institute of Informatics of the HES-SO Valais Wallis, has secured a very ambitious Spark grant from the Swiss National Research Foundation. Working in the Human-Centered Computing Group, he is interested in human-machine interactions. At a time when questions are being asked about the replacement of humans by artificial intelligence, this type of research seems essential.
Automated vehicles have been hailed as a technology with the potential to bring about profound changes in transport and the organisation of urban life. Between 2016 and 2022, according to the Federal Roads Office, ten pilot tests of autonomous vehicles for public transport were carried out in Switzerland. Although the pilot tests yield very useful results, reports and publications usually provide little detail on the practical knowledge and skills developed by the professional participants. Indeed, the operators and technicians who carried out the tests on the street solved technical and social problems and guided passengers and others on the public highway in their interactions with shuttles.
The Spark project seeks to fill the identified gaps and explore the details of these informal knowledge and skills. Moving from an implicit resource to an explicit subject, this knowledge will be obtained retrospectively from former participants in autonomous vehicle trials. By collecting the “oral histories” of these trials, the aim is to explore the social implications of automated mobility and AI by understanding how former participants account for the organisation of the pilot trials and their own involvement in the trials.
Thanks to numerous video recordings made during the tests, the social interactions between the operators and the passengers are documented. These recordings also show that the independence of vehicles is quite relative, since the operators do a great deal of work to explain clearly to the users how the vehicles work. Interviews with the technical teams, engineers and operators will be scheduled to compile the practical and social knowledge gained during the tests carried out a few years ago. Their memory will be revived by viewing excerpts from the recordings, and their experience will enrich the social history of the technologies in order, for example, to compare the progress made. This innovative methodology, which draws on experiences of five years or more, allows us to approach the research question from a reflective and narrative point of view. The purpose of this oral history is to analyze the past and understand how that experience shaped the present identity and relationship to technologies of the interviewees. Spark funds are intended for less conventional and riskier projects. Reflecting on the present implications of past interactions between humans and machines allows us to imagine how we can do better in the future. Indeed, the name chosen for this project is Remembering the Future.
The aim of this research project is not technological, but human and social. It is a question of safeguarding field knowledge before it is forgotten, making it explicit and available to all stakeholders. It will be a question of understanding how these autonomous vehicle tests are conducted by operators in terms of work routines. “Insider knowledge” has been acquired by humans throughout the trial but has certainly not been included in the official reports. Understanding how this knowledge has been developed and transformed, and how it has been passed on and taught to the new recruits for the trials is crucial. This will help formalize knowledge that could benefit others. Thanks in particular to the SAAM (Swiss Association for Autonomous Mobility) network and the public transport operators Bernmobil and PostAuto, among others, the results will be disseminated and used.