MERCURY AI
From post-flight analysis engine to airborne cognitive system. Real-time 3D reconstruction, edge inference, swarm coordination and predictive environmental modeling — thinking at thirty thousand feet.
Compute born for flight.
Mercury runs on a custom aerospace-grade compute module, designed in-house for the thermal, vibration and power constraints of sustained autonomous flight — not adapted from ground hardware. Sealed, vibration-isolated, and rated for continuous operation across the full Phantom flight envelope.
From pixels to understanding.
Seeing While Flying
Traditional structure-from-motion needs global optimisation after landing. Mercury's hierarchical approximation builds the world frame-by-frame, live, at altitude.
Single-frame detection precision from a YOLOv26 backbone, custom-compiled for the Phantom compute module and fine-tuned for environmental targets.
Multi-frame temporal fusion lifts precision to near-certainty — stable identification across motion, occlusion and changing light.
One mind, many bodies.
Multiple UAV with a shared world model, market-based task allocation and Byzantine-fault-tolerant consensus. The swarm reorganises around failure and divides the map without a central point of control. Drag the field — push the nodes and watch them re-settle.
From observation to anticipation.
Not just data — recommendations. Mercury models fire spread, pollution transport and infrastructure degradation, delivering time-stamped action guidance while humans keep final authority.
Spread Modeling
Coupled atmosphere-fire model. Inputs: fire perimeter, fuel type, terrain slope, wind field. Output: spread probability map, time-of-arrival contours, firebreak placement.
Plume Transport
Gaussian plume model with AURA-measured wind profiles. Source location from detection, emission rate from concentration, downwind impact for evacuation planning.
Degradation Forecast
Trend analysis from repeat surveys. Erosion rate from 3D change detection, pipeline exposure trajectory, failure-probability timeline for predictive maintenance.
“Deploy containment team to coordinates X, Y. Evacuate zone Z within 45 minutes. Inspect pipeline segment P within 72 hours.” — operator approves, modifies, or overrides.
The path to autonomy.
Post-Flight Engine
Offline 3D reconstruction and detection pipeline — proving the perception stack on recorded missions.
Edge Inference
YOLOv26 on the Phantom compute module — live detection and temporal fusion in the air.
Swarm Cognition
Shared world model and consensus across multiple UAV — coordinated coverage at scale.
Predictive Authority
Anticipatory modeling with time-stamped recommendations — human-in-the-loop decision support.