University space utilisation: how to measure it, and the platforms that do it

UK universities run some of the most expensive estates in the public sector, and most of it is half-empty most of the time. Lecture theatres sit dark between bookings. Libraries are full in week 10 and quiet on a Friday in week 3. The pressure to cut cost, hit carbon targets, and still give students the space they need all lands on the same question: how is the estate actually used?

Booking data cannot answer it. A timetable shows what was scheduled, not what happened. This guide is for estates, timetabling, and space-planning teams who need the real picture. It covers the platforms universities use, the metrics that matter, and why the number you act on should come from measured occupancy, not assumptions.

The four ways universities measure space

Most institutions end up with a mix of these. Each does one job well, and each leaves a gap.

Timetabling and scheduling systems. Tools like Scientia, Syllabus Plus, CELCAT, and UniTime hold the plan: which room each class is in, how many students are enrolled, and how many hours each room is booked. They are the system of record for what should happen. They cannot tell you whether anyone turned up.

Space management and IWMS platforms. Integrated workplace management systems such as Planon, Archibus, and Concerto hold the room inventory, ownership, floor plans, and maintenance. They treat space as an asset and support allocation and capital planning. Like scheduling tools, they work from plans and records, not live occupancy.

Occupancy sensors. Sensors measure real use: how many people are in a space, when, and for how long. This is the layer that verifies the other two. It is the only source that shows a booked room nobody used, or a room in heavy use with no booking at all.

Business intelligence. Power BI, Tableau, and similar tools combine the others into dashboards for leadership. A BI layer is only as good as the data feeding it, which is why the sensor layer matters: it is the difference between a dashboard built on bookings and one built on reality.

The metrics that matter

Before choosing a platform, it helps to be clear on what these reports actually measure. Four metrics carry most of the weight.

Room utilisation is the share of available, operating hours that a room is in active use. It is the baseline efficiency measure for an estate.

Capacity ratio, or how full a room is when occupied, is headcount divided by capacity, averaged across the hours the room is in use. Utilisation tells you a room is used; the capacity ratio tells you whether it is the right size.

Timetable no-show rate is the share of booked sessions with no measurable attendance. Each no-show holds a room unavailable to others, and still costs energy, for an event that did not happen.

Lecture retention is how much of a recurring series’ opening attendance is still there by the end of term. It is a proxy for sustained engagement, and an early warning when a module is drifting.

Seat hours, peak occupancy, and turnover time round out the set, but those four answer most estates questions.

Planned versus actual: the gap booking data hides

This is the heart of it. Booking data assumes every timetabled session runs and every seat fills. A manual utilisation survey, run for a week every few years, captures a single snapshot that is out of date by the next term. Both overstate use, and the gap is usually larger than people expect.

In a study of 186 rooms across a multi-campus UK university, measured with continuous occupancy counting:

• rooms were in active use for just 39.8% of operating hours: six in ten rooms sat empty in a typical hour
• where rooms were in use, headcount averaged only 35.8% of capacity, so the occupied rooms were also around two-thirds empty
• 17.5% of timetabled events, almost one in six, recorded no attendance at all
• the same data revealed more than 25,000 off-timetable room-hours: rooms in real use with no booking behind them, activity the institution previously could not see

Median lecture retention across recurring series was 82.8%, so the average course held most of its audience to the end of term, but the spread is where the signal sits: the series losing a third of their students are visible early, well before module evaluations.

None of this is in the timetable. It only appears when you measure what actually happens.

How sensor-based occupancy counting works

There is more than one way to count people, and the method shapes cost, accuracy, and privacy.

Cameras can be accurate but need power and network cabling at every position, and they raise obvious privacy concerns in spaces full of students. Thermal sensors avoid imaging but need roughly one device per few square metres, which becomes expensive across a large estate and means a lot of hardware to install and maintain. Wi-Fi tracking reuses existing access points but is coarse and depends on capturing device identifiers. Manual surveys are cheap to start but capture only a snapshot and date quickly.

Crowd Connected uses a different approach. Battery-powered Bluetooth beacons, fixed with self-adhesive and forming a wireless mesh, passively detect the Bluetooth signals that phones and other devices already broadcast. A single gateway per building relays the data to the cloud, where a self-calibrating, machine-learning engine converts device counts into people counts. There are no cameras, no images, and no app for students to install. Because the hardware is wire-free, it works in older and listed campus buildings where cabling is difficult or impossible.

The same beacons that count occupancy also support blue-dot wayfinding and asset tracking, so a university gets three capabilities from one install rather than buying a separate system for each.

Privacy comes first

Counting students must not mean tracking them. Crowd Connected counts devices anonymously and converts them into aggregate people counts, storing no images and no device identifiers. In counting mode there is generally no personal data being processed at all, which keeps the data straightforward under UK GDPR and easy to share with students, staff, and committees. The honest version of “privacy first” is having no personal data to govern in the first place.

How to choose, and how to combine

The right tool depends on the question. If you need to know what should happen, scheduling and IWMS systems already hold it. If you need to know what did happen, you need sensors. Most universities get the best result by keeping their timetabling system, adding continuous occupancy measurement, and bringing both together so the gap between them becomes visible and actionable.

That is the point of measuring space utilisation: not a single headline percentage, but a clear, current view of which rooms and which hours are pulling the estate’s efficiency down, so you can recover room-hours, right-size teaching spaces, cut the energy spent on empty rooms, and make the case for investment with evidence rather than assumptions.

Crowd Connected provides occupancy sensing for universities including the London School of Hygiene & Tropical Medicine, Buckinghamshire New University, Oaklands College, and Oulu University. To see how the data is captured and reported, read about our occupancy analytics and space utilisation platform, or how it fits the wider university estate.