DEEPEaRth Lab develops AI-powered edge-computing systems embedded on drones to support real-time, informed decision-making during disasters and humanitarian emergencies. By extracting actionable information directly on-board, the system reduces decision latency where connectivity, time, and data availability are critical constraints.
Traditional EO
- Post-event analysis
- Cloud-dependent workflows
- Mapping-oriented outputs
DEEPEaRth Lab
- In-event decision support
- On-board edge AI processing
- Action-oriented outputs
System Architecture
Edge AI on drones
Adaptable lightweight CNNs process data directly on-board drones.
Decision Outputs
Tracking event evolution with minute-scale response times.
Responders
Civil protection and humanitarian operators receive actionable information.
Validation & Traction
Illustrative operational scenario, demonstrating robust, fully on-board processing for rapid decision support. Independent evaluations confirm the system operational readiness for deployment in challenging, data-scarce, and low-connectivity environments, supporting timely and informed interventions at scale.
Operational Impact
Designed as a life-critical decision-support system, enabling alerts, evacuations, road closures, and prioritization of interventions within minutes rather than hours, even under low-connectivity conditions.
Reducing uncertainty and decision latency during emergencies.
Team
Marco La Salandra
Lab member / Founder
Drones, Edge computing, High-performance computing
Rosa Colacicco
Lab member / Founder
Earth Observation; Cloud-based computing; Geospatial analysis
Domenico Capolongo
Lab member / Founder
Remote Sensing; Geospatial analysis; Numerical modelling