SAFE-EDGE
Workshop on Security, Availability, and Fault-Tolerance in Edge AI Systems
Workshop Chairs
- → Dinesh Kumar Sah
- → Hossein Fotouhi
- → Kapal Dev
Detail SAFE-EDGE 01/04
Edge AI systems are increasingly deployed in critical infrastructures such as intelligent transportation systems, industrial IoT, smart grids, healthcare, and autonomous cyber-physical systems. While edge computing enables low-latency and decentralized intelligence, it also introduces new security, availability, and fault-tolerance challenges in distributed, resource-constrained, and physically exposed environments. SAFE-EDGE 2026 aims to bring together researchers and practitioners working on resilient and secure edge intelligence, with a focus on architectural, algorithmic, and system-level solutions such as adversarial machine learning at the edge, intrusion detection for distributed AI nodes, fault-tolerant inference, self-healing architectures, confidential computing, secure orchestration, and formal dependability models for AI-enabled cyber-physical systems.
Topics of interest include, but are not limited to 02/04
- Security-by-design for edge AI architectures
- Availability and reliability models for distributed edge intelligence
- Fault-tolerant and resilient inference mechanisms
- Adversarial machine learning in edge environments
- Model poisoning and integrity protection in distributed AI systems
- Intrusion detection and anomaly detection for edge AI nodes
- Secure orchestration and containerized edge deployments
- Confidential computing and trusted execution environments (TEE) at the edge
- Formal methods and dependability modeling for AI-enabled cyber-physical systems
- Self-healing and autonomous recovery mechanisms
- Resilience against network partitions and intermittent connectivity
- Secure federated and decentralized learning
- Privacy-preserving AI computation at the edge
- Risk analysis and threat modeling for edge AI systems
- Edge AI resilience in critical infrastructures (transportation, smart grids, IIoT, healthcare)
- Monitoring, auditing, and accountability mechanisms for distributed AI
- Side-channel and hardware-level attacks on edge AI devices
- Trust management and trust models in distributed edge ecosystems
Workshop Chairs 03/04
Workshop Chairs
Program Committee
Submission 04/04
Submission Guidelines
The submission guidelines valid for the workshop are the same as for the ARES conference.