At a landmark commercial tower in Canary Wharf, London, the toughest room PEAK has faced this year became its strongest proof point. Ninety-eight minutes after an on-site demo began, two of the engineers who had pushed back hardest had found a live chilled water valve fault, verified it on the ground, and put the evidence on the record themselves.
At a glance
- 98 minutes from demo kickoff to field evidence logged
- £25,053 a year estimated saving once the valve is rectified
- 20 minutes for PEAK to surface the fault after the session began
- 40% open actual valve position, against 0% commanded by the BMS
- Day one: two engineers self-serving on the platform
Challenge
Rolling out any analytics platform to an experienced engineering team comes with a familiar obstacle: scepticism. The engineers at this site knew their plant intimately, and they asked the fair question every good engineer asks — what could a platform surface that they didn't already know?
The building's engineering services provider had recently adopted PEAK to bring independent, data-driven fault detection to its maintenance operations. At 14:30 on an August afternoon, the CIM team sat down with the site engineering team for an on-site demo and training session. It was a hard room: real challenge, real scepticism, and no appetite for dashboards for their own sake.
The moment
Twenty minutes into the session, PEAK raised a live alert on one of the building's full fresh air handling units: a suspected chilled water valve leak. The rule had triggered on a telling combination — the chilled water valve commanded to 0%, while supply air was running more than 3°C below outside air. The unit was cooling when nothing was asking it to.
Two of the engineers who had challenged the platform hardest took the alert on their phones and walked straight to the plant room.
At the unit, they tested the valve actuator: 24V power supply present, and a 6.1VDC control signal measured on the 0–10V input — equating to roughly 39% open once the inverted signal was accounted for. That agreed with what they found on the valve itself: the stem sitting 40% open, despite the BMS not calling for any cooling. The alert was right.
By 16:08 — 98 minutes after the demo kicked off — the diagnosis and photographic evidence had been posted into the PEAK ticket directly from the plant room. Fault confirmed, evidence attached, record complete.
Impact
A passing chilled water valve on a full fresh air unit means the building pays to cool air that never needed cooling. Rectifying this single fault is expected to return in the order of £25,053 a year — and it was caught in hours, rather than waiting to be noticed at the next quarterly inspection, if at all.
The fault was raised as a P2 urgent energy action, confirmed on site, and moved into the remedial workflow the same afternoon.
Just as importantly, the loop closed without anyone chasing it. The platform raised the fault, the engineers verified it, and the evidence went on the record — no follow-up calls, no reliance on anyone remembering what they found in the plant room.
Why this one matters
Platform adoption doesn't come from the people who nod along in the demo. It comes from the ones who challenge it, then go and test it for themselves. The engineers who pushed back hardest were the first to put PEAK to the test — and they proved its value through their own investigation, on day one.
PEAK raises actions like this every day. Some save energy, others prevent failures — but they only deliver value when someone takes ownership and sees them through. Ninety-eight minutes from a demo room to confirmed field evidence is what that looks like when the platform and the people on the ground work together.
Results
- Live chilled water valve fault detected 20 minutes into the team's first PEAK session
- Fault independently verified in the plant room by site engineers, with actuator readings and photographic evidence
- £25,053 estimated annual saving once the valve is rectified
- Evidence logged in PEAK from the plant room 98 minutes after the demo began
- A sceptical engineering team self-serving on the platform from day one
