Abstract

Technological advances in distributed cyber-physical systems (CPS) will fundamentally alter the way present and future human societies lead their lives. From a security or privacy perspective, a (multi-agent) cyber-physical system is a network of sensors, actuators, and computation nodes, i.e., a system with multiple attack surfaces and latent exploits that originate both through software attacks and physical attacks. In this paper, we argue that we are in pressing need to bring about a paradigm shift in software development for multi-agent CPS. To this end, security and privacy policies should be made a critical ingredient of agent interfaces with a goal of ensuring both localized safety and privacy for each agent, as well as guaranteeing global system safety and security. We present our vision on new theory, algorithms, and tools to foster a culture of secure-by-design multi-agent CPS.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call