ROSS, for Remote Observation, Sensing & System, is the applied R&D team established in summer 2026 under the Multimedia Research Centre (MRC) at the University of Alberta by Professor Irene Cheng and Alvin (Xinyao) Sun. This is the story behind it, as told at the first ROSS Club event on August 19, 2026: where the work came from, what it taught the group about real-world AI, and how students and partners can join the pathway.
Remote observation at MRC
MRC is a long-standing research centre in Computing Science, led by Professors Anup Basu and Irene Cheng, with a deep record in computer vision and signal processing. It also runs the industry-driven Multimedia Master’s Program, and its lab in UCOMM 4-140 is now home to ROSS and the ROSS Club. Since 2017, MRC’s remote sensing research, led by Professor Cheng, has gradually expanded into intelligent data analysis and applied AI across agriculture, healthcare, robotics, digital twins and more.
Built through applied research
- 2017, space. Wide-area InSAR monitoring deployed with 3vGeomatics, supported by the first CARIC funding awarded in Alberta together with Mitacs. That project later grew into trust-scored monitoring.
- Airborne and ground. The same methods extended to mining, wildfire and agriculture scenarios, and to LiDAR and digital twins with McElhanney.
- Indoors as well as outdoors. The same sensing questions moved from fields and sites to rooms and clinics: health monitoring, telemedicine and rehabilitation.
- 2025, dual-use. Engagement with CARDD-Tech, the Centre for Applied Research in Defence and Dual-Use Technologies.
- Strengthened by the MM co-op program and MRC thesis research. More than 250 graduates, about $6M in grants for industry R&D projects, and an expanding partner network.
What real-world AI demands
AI will continue to grow. The challenge is using it correctly to solve real problems. Across every sensor, application, platform and technology readiness level the group has worked with, the same six challenges recur:
- Imperfect data. Noisy, sparse and non-ideal observations.
- Embedded expertise. Domain knowledge that is difficult to formalize.
- Multimodal sensing. Decisions built from multiple sensors and data types.
- Edge constraints. Limited compute, privacy, security and autonomous operation.
- Operator-ready outputs. Evidence translated into actionable decisions, not just predictions.
- Traceability. An auditable path from each decision back to the observation.
This is why the latest model cannot simply be dropped onto a real problem. ROSS treats these six challenges as shared R&D building blocks and adapts them to each application.
Why ROSS, why now
Put the applied pathway on a timeline and the reason becomes clear: 2015 MM co-op growth → 2017 wide-area monitoring (CARIC and Mitacs) → 2020 health and edge AI → 2022 to 2024 agriculture and robotics → 2025 CARDD-Tech engagement → 2026, ROSS established as a dedicated applied R&D team. By 2025 this branch of work had enough momentum to benefit from a team of its own.
ROSS builds on MRC’s foundational computer vision and signal processing, connects projects across domains, and helps promising methods move toward deployment. The team works as Forward Deployment Research: researchers work alongside partners on their real data, sites and constraints, so methods are shaped by deployment from the first day rather than adapted after the fact. Its mission fits in four words: remote signals, trusted decisions. A single pixel or sensor reading has limited value by itself. Its value comes from the decision it supports, the reasoning behind that decision and the benefit the overall system delivers to the end client.
Growing the pathway
An R&D team is half of the story. The other half is a wider pathway for students and partners:
- Learn. Remote observation is part of the curriculum through CMPUT 617, Remote Observation, taught by Professor Cheng since 2024 and by Alvin Sun in Fall 2026.
- Partner. The MM program brings industry problems onto campus, where students work on them through course projects.
- Connect. Agriculture, healthcare, engineering and space.
- Participate. Course projects, research, co-op placements and theses, which also help students find out whether industry or research is the better fit.
The ROSS Club
Partners bring more real problems and opportunities than one research team can pursue, so the group launched the ROSS Club to open those connections to students across the university. It is student-oriented, open to all UofA students and has no prerequisites. Rather than a series of information sessions, the Club is built to learn (talks, regular meetups, practical guidance), build (partner applications turned into public challenges through workshops and hack sessions, with field demonstrations and site visits when possible) and connect students with ROSS researchers, graduates and industry partners. Its first event, Remote Sensing in Agriculture with Joshua Billson, took place on August 19, 2026.
To stay involved, join the newsletter, explore the research on this site, or visit the ROSS Club page. Speakers, datasets and partnership proposals are welcome at any time.
Further reading: ROSS Club newsletter