0539640 - ÚI 2021 NL eng V - Výzkumná zpráva
Wiedermann, Jiří - van Leeuwen, J.
Autonomous Vehicles that Cooperate and Understand.
Utrecht: Utrecht University, 2021. 22 s. Technical Report, UU-PCS-2021-01. ISSN 0924-3275
Grant ostatní: AV ČR(CZ) StrategieAV21/1
Program: StrategieAV
Institucionální podpora: RVO:67985807
Klíčová slova: autonomous vehicles * driving algorithm * machine consciousness * machine understanding * philosophy of computing * safety * smart roadside infrastructure

We present a new paradigm for the research and development of autonomous vehicles from a philosophical viewpoint. The approach takes the broader epistemic context into account in which the vehicles operate and utilizes cognitive mechanisms inspired by higher-level mental processes. The goal is to design (connected) autonomous vehicles that provably understand all traffic situations that they can perceive and that cooperate with an informed smart road infrastructure in resolving them. The features are seen as a cognitive attainment – the vehicle’s ability to handle traffic situations in a way aligned with its mission, based on information about the surrounding traffic and about the past actions of the vehicle. A key ingredient of the approach is to view an autonomous vehicle as a cognitive cyber-physical human system endowed with so-called ‘minimal machine consciousness’, a prerequisite of machine understanding. Its on-board sensors and the external smart road infrastructure must provide a vehicle with the information that is sufficient to provably elicit its understanding of the evolving traffic situations and fulfil its mission, in cooperation with other vehicles and the smart road infrastructure. We show how the approach leads to a driving algorithm that is arguably safe and reliable for guiding a connected autonomous vehicle to its destination. We discuss the potential of the new paradigm to overcome the difficult issues in autonomous driving.

Trvalý link: http://hdl.handle.net/11104/0317346