Rajdeep Singh

← Back to work

AI-Enabled Safety Assurance for Swarm UAV Operations

surveyswarm roboticsfault tolerance

Survey on making AI-driven drone swarms safe enough for civilian use — covering what breaks, how to detect it, and how to test it.

Scope

  • Reliability — fault tolerance, redundancy, and maintenance modeling (k-out-of-n, FMEA) for system, component, and communication failures
  • Anomaly detection — statistical and ML methods for GPS/INS drift; IDS and ConvNet approaches for communication faults
  • Communication — direct, indirect, and ACL-based architectures; centralized vs. distributed control
  • Formation and navigation — formation control, route planning, obstacle avoidance
  • Testing — simulation frameworks and validation strategies

Applications

Surveillance, delivery, search and rescue, urban air mobility.

ArXiv link coming soon.