CO2, Humidity and Open Windows: How Elsys Sensors and Vemco Keep University Buildings Healthy and Safe

A lecture hall built for 120 students hits 1,400 ppm CO2 forty minutes into a class of 85. That is not an overcrowding problem — it is a ventilation problem, and it looks completely different on a dashboard than the same reading in a room packed with 140 people. This distinction is the whole point of pairing air quality sensors with occupancy data, and it is where most university monitoring projects quietly fail: they buy the sensors, mount them, and then stare at CO2 curves with no idea how many people were actually in the room when the line spiked.

If you manage facilities, safety or energy budgets on a campus, you have almost certainly read the generic advice by now — "CO2 above 1,000 ppm impairs concentration," "humidity should sit between 40 and 60 percent." True, and useless on its own. What you need is the operational layer: which rooms, at which hours, under which occupancy loads, and who gets told when something is wrong. That is the layer this article covers, using the way more than 100 universities run Elsys sensors through the Vemco platform as the working example.

CO2 per person, not CO2 per room

Elsys sensors measure CO2 and humidity in lecture rooms, libraries and offices. On their own, those readings tell you a room is stuffy. Combined with occupancy on the same screen — which is how the Vemco platform presents them — they tell you why it is stuffy, and that changes the fix entirely.

• High CO2, high occupancy: the room is being used beyond its ventilation design. The fix is scheduling — move the class, cap the booking capacity, or split the cohort. No mechanical work needed.
• High CO2, normal occupancy: the ventilation is underperforming. Now it is a maintenance ticket — a failed damper, a miscommissioned AHU schedule, a filter overdue for replacement.
• Low CO2, low occupancy, full ventilation: you are conditioning air for nobody. This is where energy managers find their quickest wins — often in libraries and seminar rooms that run full HVAC schedules across reading weeks when footfall drops to a fraction of term-time levels.

Facility teams that only see one half of this equation end up either over-ventilating everything as a precaution — expensive — or fielding complaints from staff and students with no data to triage them. Seeing both halves is what turns indoor air quality monitoring for universities from a compliance checkbox into a scheduling and energy tool.

Humidity: the slow-moving budget risk

CO2 gets the attention because people feel it. Humidity is quieter and more expensive. Sustained readings above 60 percent RH in older campus buildings — and most universities own plenty of pre-1980 stock — create the conditions for mould in libraries, archives and basement teaching spaces. Sustained readings below 30 percent in winter dry out timber, aggravate respiratory complaints and generate the static-and-nosebleeds emails your helpdesk knows well.

The value of continuous humidity data from Elsys sensors is not the reading itself but the trend. A library stack that creeps from 55 to 68 percent RH over three weeks after a roofing repair is a problem you catch for the price of a maintenance visit. The same problem discovered six months later, when a special collections librarian finds bloom on book spines, is a remediation project with a different number of zeros on it. Health and safety officers should also note that humidity trend logs are increasingly requested in workplace complaint investigations — having twelve months of timestamped data ends arguments quickly.

The open window problem, solved specifically

Every campus security team has a version of the evening walk-round: someone with a torch checking hundreds of windows across dozens of buildings because one open casement means overnight heat loss, rain damage, or an intrusion route. It is hours of labour to find, on most nights, nothing.

The Vemco platform can identify which specific window was not closed after closing time, so staff go straight to Building C, second floor, room 214 — instead of walking every corridor on the off-chance. On a large campus this is the difference between a two-minute task and a two-hour patrol, and in winter it is measurable on the heating bill: an open window feeding cold air onto a thermostat can drive a whole zone's radiators at full output all night.

Delivery matters as much as detection. Alerts from the platform go out via app, email, WhatsApp, SMS, webhook or MQTT — which sounds like a feature list until you realise it means the security team on evening shift gets a WhatsApp message, the BMS gets an MQTT payload it can act on automatically, and the facilities helpdesk gets a ticket via webhook, all from the same event. Nobody has to log into a dashboard at 22:40 to find out something is wrong.

What an implementer would tell you before you start

Here is the observation that rarely makes it into vendor material: sensor placement kills more campus IAQ projects than sensor quality does. A CO2 sensor mounted near a doorway or under a supply diffuser will read fresh air all day while the back rows of the lecture hall sit at 1,600 ppm. Mount at breathing height, away from windows, doors and vents, and — critically — walk the room with a handheld meter during a real lecture before you finalise positions. Budget half a day per building for this. Teams that skip it spend the first term explaining why the data contradicts what occupants are complaining about, and the project loses credibility it never fully recovers.

Two procurement notes worth having in writing before you sign anything. First, Vemco is sensor-agnostic — Elsys is the common pairing on campuses, but if your estates team already has LoRaWAN devices deployed, they do not become stranded assets. Second, the platform runs on AWS in EU Frankfurt, which shortens the data protection conversation with your DPO considerably compared with platforms hosted outside the EU. Occupancy data from teaching spaces attracts GDPR scrutiny even when it is anonymous counting; hosting location is one of the first questions you will be asked.

On counting accuracy, since occupancy underpins the per-person air quality analysis: the honest figure is a contractual minimum of 96 percent, typically reaching 98–99 percent where conditions such as lighting, layout and visitor behaviour allow. Be wary of anyone quoting a flat guaranteed figure without those caveats — atria with heavy glazing and doorways where students cluster are genuinely harder counting environments, and a supplier who acknowledges that is telling you the truth about the rest of their claims too.

Where this leads: the right space at the right time

Once environment data and footfall-based occupancy sit in the same system, you have the foundation for the bigger question every university estates strategy is now asking: are we allocating the right space at the right time? Which buildings can close early on Fridays, which lecture halls are chronically over- or under-booked relative to their ventilation capacity — those decisions become evidence-based rather than anecdotal. That is a larger topic than this article, but the sensors described here are where it starts.

If you are weighing up CO2 and humidity monitoring for your campus, or you want to stop paying for nightly window patrols, talk to the team that has done this with more than 100 universities. Contact Vemco for a walkthrough of how Elsys sensors and occupancy data would map onto your specific buildings — bring your worst lecture hall and your longest security patrol route, and start there.