Cyberphysical risks of hacked internet-connected vehicles

Area of Science:

• Cyberphysical systems
• Traffic flow dynamics
• Network science

Background:

• Internet-connected vehicles integrate automotive technology with digital connectivity, introducing novel cybersecurity risks.
• Unlike traditional data, these vehicles are cyberphysical systems with real-world environmental interactions.
• Cybersecurity threats to vehicles can lead to emergent phenomena in traffic flow.

Purpose of the Study:

• To investigate the impact of mass, simultaneous vehicle disabling cyberattacks on traffic.
• To explore the applicability of percolation theory in predicting post-hack traffic conditions.
• To assess the vulnerability of urban street networks to such cyberphysical attacks.

Main Methods:

• Agent-based simulations were employed to model post-hack traffic scenarios.
• An analytic percolation-based model was developed to estimate road conditions.
• The model was applied to the Manhattan street network for vulnerability assessment.

Main Results:

• Percolation theory is critically relevant for probabilistically predicting traffic outcomes after vehicle-targeted cyberattacks.
• A percolation-based model can rapidly assess road conditions based on the density of disabled vehicles.
• The study quantifies citywide traffic disruptions, estimating the critical number of disabled vehicles impacting traffic flow.

Conclusions:

• The statistical physics of percolation offers a method to estimate traffic disruption from cyberphysical attacks on vehicles.
• Urban environments like Manhattan are vulnerable to cyberphysical attacks leading to widespread traffic disruption.
• This research provides an upper-bound estimate for traffic disruptions caused by multiple hacked vehicles.