Organisations collect vast amounts of data about their buildings – but far less insight into how people actually use them. At La Trobe University, Innovation Central Melbourne is bridging that gap by combining existing platforms such as Cisco Spaces with computer vision and AI to turn space usage into evidence-based design capability.
Most organisations can tell you how big their buildings are. Far fewer can say how well those spaces are actually working.
Across offices, campuses, and public buildings, decisions about layout, refurbishment, and utilisation are often based on assumptions: how many people should be there, how long they might stay, how spaces were intended to be used. What’s missing is reliable, real-world evidence about what actually happens once people move through the door.
That gap is what drew Innovation Central Melbourne (ICM) into the challenge of building an occupancy and utilisation capability.
Working with existing enterprise platforms already deployed in buildings – including occupancy and location systems such as Cisco Spaces – ICM’s team saw an opportunity to go further. Not by replacing core infrastructure, but by extending it. The question wasn’t how to collect more data, but how to extract better insight from what was already there.
“Most buildings are already generating signals,” says Jeff Jones, Director of ICM. “The challenge is turning those signals into something decision-makers can use.”
Turning signals into behaviour
The project focused on layering new capability over an existing sensing environment. Rather than relying on raw counts or static heatmaps, ICM explored how computer vision and AI could add a higher-order understanding of how people behave in shared spaces – while still operating within strict privacy and metadata constraints.
A key contribution came through an industry-embedded internship with Minh Duc Truong, who designed and built a computer-vision pipeline that works alongside existing sensor data. The system can identify groups of people, estimate group size, track how long they stay, and observe how occupants move between zones – without identifying individuals.
This additional layer shifts the question from how many people are present to how spaces are actually being used. Meeting rooms, for example, can be assessed not just on booking data but on how often groups form, how large they are, and whether the space supports collaboration or dispersal. Transitional areas can be analysed for congestion or underuse. Informal zones can be evaluated against their design intent.
“Instead of guessing whether a space is fit for purpose, you can start to measure it,” says Mr Jones. “That changes how you design, manage, and justify investment in buildings.”
From research to real-world decisions
This project also demonstrates ICM’s broader capability: taking established platforms and extending them through applied engineering, AI, and user-focused design. Rather than treating technology as a finished product, the team treats it as a foundation that can be adapted, tested, and improved in real environments.
The project also highlights La Trobe’s role as a living laboratory, where campus-scale complexity allows new ideas to be trialled under realistic conditions, revealing what works, what doesn’t, and what needs refinement before broader deployment.
For ICM, occupancy and utilisation is not just a technical exercise, it’s a case study in how universities can help organisations move from invention to innovation. “Universities are uniquely placed to do this kind of work,” says Mr Jones. “We can experiment safely, work across disciplines, and focus on outcomes rather than products.”
