Realising Meaningful Human Control Over Automated Driving Systems: A Multidisciplinary Approach.

Area of Science:

• Human-Computer Interaction
• Autonomous Systems
• Philosophy of Technology

Background:

• Automated Driving Systems (ADS) present challenges in maintaining human control and accountability.
• Existing systems risk creating responsibility gaps in safety-critical operations.
• Multidisciplinary research is needed to define and ensure meaningful human control.

Purpose of the Study:

• To present a novel framework for achieving meaningful human control over ADS.
• To operationalize the concepts of tracking and tracing for ADS control.
• To address safety concerns and reduce responsibility gaps in automated driving.

Main Methods:

• Synthesized findings from a multidisciplinary research project (2017-2021).
• Developed a framework based on tracking (ADS behavior aligns with human reasons) and tracing (events link to human actors).
• Operationalized conditions using a proximal scale of reasons and an evaluation cascade table.

Main Results:

• Proposed a framework where ADS are under meaningful human control if they track relevant human reasons and trace dangerous events to human actors.
• Operationalized tracking and tracing conditions for practical application.
• Identified implications for human actor behavior, skills, and driver education.

Conclusions:

• The framework provides a method to ensure human control and moral responsibility in ADS.
• The evaluation cascade table aids in identifying traceability deficiencies in engineering use cases.
• Further research is needed for framework expansion, real-world pilots, and institutional embedding.