Observing the WORLD and benefiting HUMANITY, remotely.
ROSS — Remote Observation, Sensing & System
Any sensor · Any platform · Space-borne · Airborne · Robotic
AI, signal processing and communication turn remote signals into decisions, from research to real deployments.
What we watch.
All 6 themes →Earth Observation
Wide-area ground monitoring from space. Satellite radar becomes trust-scored movement signals, watching roads, rail, dams, mines and critical infrastructure.
Agriculture & Environment
From pasture health and rice mapping to wildfire risk. AI on satellite and drone sensing in service of food, land and environment.
Autonomy & Perception
Multi-modal perception for platforms that sense on the move. Thermal-RGB fusion for detection and monitoring, from satellite to drone to robot.
Dual-Use & Defence
Civilian sensing carried into defence and dual-use contexts, in two streams. Dual-use sensing on one side, and cybersecurity with blockchain-secured data systems on the other.
Health & Tele-medicine
Carrying the group's imaging and AI lineage into medical image analysis and remote, tele-medical sensing: denoising, segmentation, multi-modal fusion, and learning from scarce labels.
Digital Twin & Simulation
Synthetic worlds that train and test sensing AI. InSAR simulators, generative scene data and edge-ready models, moving toward living digital twins of monitored environments.

An AI denoiser with a per-pixel trust score
Clients don't just get a movement map: they get movement plus a 0–1 trust score for every spot. AI cleaning cut new-site turnaround from ~1 month to a single InSAR pair.
Read the case →
MatrixShield: adversarial security testing for AI agents
A dual-use security showcase: MatrixShield red-teams autonomous AI agents as a real adversary would, grading every response with an independent judge to turn agent risk into a board-ready 0-to-100 score. A collaboration led by Alvin (Xinyao) Sun with Matrix Labs.
Read the case →
Large-scale rice mapping with vision transformers
Vision transformers on multi-temporal Sentinel-1 SAR time series map rice fields across nine regions of Brazil — 42,481 km² — outperforming LSTM, TFBS, SegFormer and SETR.
Read the case →
Tessera: a smart remote-observation platform, from cloud to orbit
A smart remote-observation platform by MM2025 student Shuping Tan: draw an area of interest, pick an algorithm, and get a biomass, methane or marine-plastic map in minutes. Co-developed by MRC / University of Alberta and Matrix Labs.
Read the case →
BAAS UA-MM Consortium: verifiable academic records on Hyperledger Fabric
A four-organization Hyperledger Fabric consortium built with MatrixLabs for the UofA Multimedia Master's program: role-scoped portals, one-signature verification, and privacy-preserving records.
Read the case →
Drone-based cattle monitoring and weight prediction
A thermal-RGB fusion pipeline detects cattle from drones in field work with LandView and Serecon, and a Three-Stream DCNN predicts cattle weight from single-view images (example output 294 kg).
Read the case →Research that leaves the lab
Latest from the group
All articles →ROSS Club First Event: Remote Sensing in Agriculture, by Joshua Billson
Introducing ROSS: a decade of watching the ground from space
How a machine reads an image: features, histograms and colour scales
From space to street: flood assessment on public map data
The four levels of remote observation: space, air, land and sea
Why observe Earth from space? Canada's case for EO
Learn the field.
All articles →Short, informative reads on the ideas and methods behind the work, across every theme: from satellites to robots to the lab.






