Investigating Car Drivers’ Information Demand after Safety and Security Critical Incidents

Title of the Paper

Investigating Car Drivers’ Information Demand after Safety and Security Critical Incidents

Paper Information

• Research Domain: Human-Computer Interaction, Automotive Automation, Risk Perception
• Keywords: Information Demand, Automotive Safety, User Satisfaction, Technical Failures, Malicious Intrusions

Research Background and Problem

• Issues and Challenges:
  • As modern vehicles become increasingly automated, the sources of car malfunctions are becoming harder for drivers to understand (e.g., technical failures and malicious intrusions).
  • After encountering safety and security-critical incidents, drivers require additional information to comprehend vehicle behavior. However, it remains unclear which types of information are most important and helpful to drivers.
• Significance of the Problem:
  • Improved information provision can not only help drivers understand vehicle behavior and build trust in technology but also clarify accountability in case of accidents and reduce the likelihood of similar incidents in the future.
• Related Research:
  • Previous studies on information demand in context-aware applications and web security warnings have primarily focused on low-risk scenarios and have not addressed high-risk, highly automated contexts such as automotive systems.

Solution

• Research Methodology:
  • A mixed-method online survey was conducted, featuring two study scenarios: an autonomous vehicle colliding with a construction barrier on a highway (high risk) and a remote key failure (low risk).
  • Three types of explanations were provided in the scenarios: Malicious Intrusion (MI), Technical Malfunction (TM), and No Explanation (NO).
  • The survey assessed participants' understanding of the scenarios, information demand, trust levels, satisfaction, and subsequent behavioral intentions.
• Innovative Contributions:
  • Systematically revealed drivers’ information needs in different critical incidents, providing data to design context-adaptive communication between vehicles and drivers.
• Technical Methods and Implementation Steps:
  1. Designed an online questionnaire containing both qualitative and quantitative questions.
  2. Applied Thematic Analysis to code and analyze qualitative data, combined with Correspondence Analysis to explore how contextual factors influence information demand.
  3. Evaluated the impact of scenario context and explanation type on drivers’ specific information demand characteristics.

Research Findings

• Key Discoveries:
  • Drivers consistently demanded basic information across all scenarios, including “What happened,” “Why it happened,” and warning messages.
  • Drivers exhibited low awareness of malicious intrusions, typically considering security threats only after being prompted.
  • Explanations related to technical malfunctions often led drivers to focus more on internal vehicle state information, whereas malicious intrusion explanations increased interest in attack details and event context.
• Experimental and Evaluation Results:
  • High-risk scenarios (e.g., collisions) elicited broader information demands, but participants’ trust and satisfaction levels were significantly lower compared to low-risk scenarios.
  • When no explanation was provided, drivers tended to simplify the cause as either technical failure or human error.
• Limitations and Future Directions:
  • The survey sample was skewed toward technology-sensitive groups (e.g., participants primarily sourced from MTurk), potentially introducing sample bias.
  • Future research could focus on real driving contexts to explore information demand, using driving simulation experiments or testing with vehicles featuring higher levels of automation to assess practical applicability.

Recommended Designs:

1. Provide Specific Actionable Advice: Push clear next-step recommendations based on different scenarios (e.g., contact police or manufacturers).
2. Communicate Threats Explicitly: Deliver timely and precise threat notifications in critical situations.
3. Explain Preventive Measures: Inform drivers on how to avoid similar issues in the future.
4. Disclose Attack Information: For security incidents, provide drivers with moderate details about the attack and investigative clues.
5. Enhance Alert Mechanisms: Offer clear visual and auditory warning signals to draw drivers’ attention to potential malfunctions.

These design recommendations aim to enhance user trust in autonomous driving features and improve system response transparency, thereby reducing safety risks caused by insufficient information provision.