Human-Machine Bidirectional Trust-Aware Analysis and Design for Human-Led Truck Platooning
Human-led truck platooning, where a human-driven truck leads one or more autonomous followers, offers significant benefits in fuel efficiency, safety, and traffic flow. However, its successful deployment hinges on trust between the human driver and the automated systems. Unlike conventional automation, trust in this context is inherently bidirectional: the human must trust the autonomous followers, and the followers must reliably interpret and respond to the human's behavior. While prior research has extensively studied human trust in automation, the reciprocal nature of trust, especially considering the expertise of professional truck drivers, remains underexplored. This paper develops a conceptual framework of bidirectional trust for human-led platooning systems. Drawing on established trust theories (ability, benevolence, integrity) and insights from truck driver psychology, we propose distinct dimensions for human-to-automation trust and automation-to-human trust. To move beyond conceptualization, we introduce a quantitative model that operationalizes the bidirectional dynamics, using the following distance as the key interaction variable to illustrate how trust co-evolves through a feedback loop. Simulation examples demonstrate both positive reinforcement and negative spiral effects. Based on this framework and its quantitative instantiation, we derive design guidelines for autonomous followers to foster appropriate trust calibration, improve safety, and enhance user acceptance. The framework bridges human factors and engineering perspectives, providing a theoretical and preliminary quantitative foundation for future empirical and modeling research.