Training Description
Key Takeaways
- Workshop on security, availability, and fault-tolerance in edge AI systems
- Focus on resilient and secure edge intelligence for critical infrastructures
- Addresses challenges in distributed, resource-constrained environments
- Brings together researchers and practitioners from academia and industry
- Explores architectural, algorithmic, and system-level solutions
SAFE-EDGE is a specialized workshop dedicated to advancing the security, availability, and fault-tolerance of edge AI systems. Held in conjunction with the 21st International Conference on Availability, Reliability and Security (ARES 2026) in Linköping, Sweden, this event gathers experts to address the unique challenges faced by edge intelligence in critical infrastructure sectors.
Overview of SAFE-EDGE Workshop
The SAFE-EDGE workshop is designed for professionals and researchers working at the intersection of artificial intelligence, edge computing, and cybersecurity. The event focuses on the deployment of edge AI systems in environments such as intelligent transportation, industrial IoT, smart grids, and healthcare. These sectors demand robust solutions to ensure that edge AI remains secure, available, and resilient despite the constraints and exposures inherent to distributed systems.
Participants will engage in discussions and presentations that highlight the latest research and practical approaches to overcoming technical and operational barriers. The workshop emphasizes the importance of resilient and secure edge intelligence, particularly in physically exposed and resource-limited settings.
Core Topics and Themes
SAFE-EDGE covers a comprehensive range of topics relevant to edge AI security. Key areas include security-by-design for edge AI, availability and reliability models, and fault-tolerant inference. The workshop also explores adversarial machine learning, model poisoning, intrusion and anomaly detection, and secure orchestration.
Additional themes involve confidential computing, formal dependability modeling, self-healing mechanisms, resilience to network partitions, secure federated learning, privacy-preserving computation, risk analysis, monitoring, accountability, side-channel attacks, and trust management in distributed edge ecosystems. These subjects are critical for ensuring the dependability and trustworthiness of AI-enabled cyber-physical systems.
Audience and Value Proposition
The workshop is tailored for researchers, practitioners, and professionals from both academia and industry. Attendees typically include security architects, AI and machine learning engineers, system designers, reliability engineers, and technical leads from sectors such as transportation, industrial automation, healthcare, and smart grids.
SAFE-EDGE provides a platform for knowledge exchange, networking, and community building. The event encourages the sharing of research findings and practical solutions, fostering thought leadership in the field of edge AI security. By focusing on education and collaboration, the workshop supports the development of resilient and secure edge intelligence for critical infrastructures.