Using AI methods, we are developing an attack tree generator that automatically enumerates cyberattack vectors for industrial control systems in critical infrastructure (electric grids, water networks and transportation systems). The generator can quickly assess cyber risk for a system at scale.

Our team will approach this problem by developing an AI planning system that can enumerate a set of multi-step attack plans capable of penetrating and compromising systems in the selected critical urban infrastructure sectors. Dr. Howard Shrobe’s work on Computational Vulnerability Analysis for Information Survivability will be used as the core of the attack graph generator; its ontology and knowledge base will be enhanced to reflect today’s urban cyber infrastructure. The attack graphs developed by this planer will provide automatic identification of concrete adversarial strategies aimed at compromising transportation systems and water networks. The attack vectors will be prioritized based on Gregory Falco et al.’s research SCADA Risk Modeling for Critical Infrastructure Cybersecurity in Smart Cities. In addition to developing an automated attack generator, the team will also develop a counter-planning system that will generate countermeasures and mitigation strategies. These will consider multi-prong attack scenarios where multiple attack vectors are pursued to compromise a city-wide sector. The counter-measures will be ranked both by coverage (number of attack plans prevented) and by cost (difficulty of implementation). Together the set of attack plans and their counter-measures will provide insight to the operators of urban critical infrastructure, illustrating worst-case scenarios and enabling an assessment of cyber risk. By having an understanding of prioritized adversarial actions and appropriate countermeasures, the team will explore how local policy can be crafted to help secure critical urban infrastructure against the most pressing security threats.

Research Areas

Impact Areas