On-board operators' workplaces in real-world laboratories

Abstract

Background: Highly automated vehicles (HAVs) at Level 4 (SAE, 2021) bear the potential to make mobility safer, eco-friendlier, and more inclusive. Across Germany, a plethora of real-world laboratories with HAVs providing novel mobility solutions in real-world settings has been implemented recently. Due to technological and legal restraints, a steward is required on board of the HAV in all of them. This on-board operator’s task is overseeing driving operations and intervening if necessary, i.e., when the automation’s capability does not suffice to maintain safe driving operations. Method: Three major HAV projects are presented as case studies with a focus on the on-board operators’ workplaces, their tasks, scenarios they are exposed to, and interactions with the driving automation to resolve these scenarios. After analyzing the similarities and differences, structured interviews with on-board operators were conducted. The interviews focused on the operators’ working environments, their tasks, and scenarios they were exposed to. Additionally, observations of HAV operations in the shuttles they supervise were carried out. The observations examined situations in which operators intervened, their frequency, and the ratio of automated versus manual driving periods. Results: Results showed a highly similar task structure of on-board operators across the three investigated projects despite considerable differences regarding the HAVs’ capabilities of automation, frequency of manual interventions, and the operational contexts. For instance, the number of averaged manual interventions per hour varied between 6 and 30 across the different operational environments in the three real-world labs. In the same vein, the percentage of driving time spent in the manual mode among the overall driving time differed considerably between the environments, from 6 to 18 percent of manual driving. Moreover, the setups of workplaces varied substantially, particularly regarding the transparency of the automation’s decision-making process and ways to interact with the automation. Conclusions: Results highlight the importance of on-board operators’ workplace design for operating HAVs safely and efficiently. In accordance with Human Factors guidelines, implications for improved workplace design and further research will be discussed.