Electrical faults trip. Cooling degrades, and degradation is a measurable signal weeks out in telemetry you already collect. That warning window is the whole opportunity.
A 450,000 sq ft, 109 MW campus. Listed among the Uptime Institute's ten example Category 5 outages of 2025, and the only cooling failure on that list.
Half of all major outages clear six figures and one in five clears seven, drawn from operators reporting on their own most recent incident.
Of most recent major outages: 43% under $100K, 37% between $100K and $1M, and 20% above. At that level a single prevented failure pays for years of monitoring.
Among mid-size and large enterprises, with 41% putting a single hour between $1M and $5M or more. That is one organization's loss, not a whole facility.
The famous "$9,000 a minute" traces to a vendor-sponsored study of 63 US sites and has not been refreshed in a decade. If a vendor quotes it to you this year, they have not checked.
Counting outages makes cooling look like the second problem. Counting what insurers actually pay out makes it the first.
Cooling has held a 13% to 19% band across recent years, second only to power. Power's share fell to 45% from 54%, so cooling's relative weight is rising.
By incident count
Across 221 data center claims worth roughly $782M, water damage led at 21%, ahead of willful acts at 19% and fire at 14%. Fire drives severity; water drives frequency.
By claim frequency
Direct liquid cooling accounts for nearly a quarter of total data center loss costs, before the technology has even finished arriving.
By loss cost
As heat transfer fades, the control system quietly raises flow to hold temperature. Outputs look fine for months while the loop degrades underneath. Then the pumps hit their ceiling and temperature runs away with no usable warning.
Watching effort instead of output is what turns a months-long hidden decline into a flagged, plannable event. That is the signature PRISM already detects in water.
For an oil-filled transformer and a main switchboard, the documented failure modes split three ways: readable from data you already stream, needs one added sensor, or genuinely needs a person. That split is the scoping conversation.
| Failure mode | Asset | Coverage |
|---|---|---|
| Cooling system failure | Transformer | Existing data |
| Overload and overheating | Transformer | Existing data |
| Insulation ageing | Transformer | Existing data |
| Partial discharge, arcing | Transformer | Added sensor |
| Oil degradation | Transformer | Lab |
| Shorted turns, winding deformation | Transformer | Outage |
| Protection setting drift | Switchboard | Existing data |
| Harmonic distortion, power quality | Switchboard | Existing data |
| Breaker duty and contact wear | Switchboard | Existing data |
| Loose and high-resistance connections | Switchboard | Added sensor |
| Protection calibration, mechanism | Switchboard | Truck roll |
| Insulation and contact resistance | Switchboard | Outage |
Load current you already have. Temperature you do not. Loose connections are the most frequent electrical failure mode and fall just outside existing data. An annual infrared scan is a snapshot at whatever load happened to be running; a joint that would be critical at full load can read clean at thirty percent. One temperature input per critical joint turns that into a continuous, load-normalized trend. It is the highest-leverage sensor addition in the asset class.
This coverage work is designed, not shipped. The classification is our engineering assessment against published reliability libraries. We do not yet have these warnings running, and we will walk through the split asset by asset rather than claim it as a live capability.
We will not hand you a dollars-per-minute rate, because the honest ones do not exist and the famous one is a decade stale. You know what an hour costs you. Put it in.
The second half is the part nobody counts: asset-level failures caught before they escalate. In the validated water deployment that was worth $1,616 per monitored asset per year, in an environment where a failure costs far less than it does on your floor. We use it as a floor, not a target.
Downtime you avoid, at your own hourly cost, plus maintenance you avoid at the rate we measured.
Basis. Downtime exposure is entirely your input. Presets anchored to published survey bands: 91% of mid-size and large enterprises put an hour above $300K, 41% between $1M and $5M or more. Maintenance avoidance at $1,616 per monitored asset per year, carried unchanged from the water deployment as a conservative floor. Excludes secondary damage, expedited freight and capacity loss from equipment lead times.
PRISM reads the BMS, CDU, EPMS and DCIM telemetry you already collect, under a read-only credential, and delivers the decision into the system your team already runs.
One-directional ingest in, decisions out. Nothing in the product can change a setpoint, close a valve or start a machine.
Your cooling plant is motor-driven pumps and fans moving a fluid around a closed loop under a control system that compensates as performance fades. That is the asset class PRISM has the deepest production history on, and your instrumentation is denser and faster than a rural lift station on a radio link.
Real failures, real dates, and the telemetry leading up to each one, across 391 assets over nine months. The method, the two-layer architecture and the tolerance for messy data all carry across unchanged.
They are the same mechanical devices at different scale, so we expect meaningful transfer. We do not expect it to transfer perfectly, and we will not tell you it does. Closing that gap is what the evaluation is for, and it runs on your data at no cost.
We will not quote a savings figure for your site until it comes out of your own history. The per-asset rate on this page was measured on someone else's fleet. It anchors the conversation; the retrospective makes it yours.
Designed to clear your security review by not needing it. Nothing connects, nothing installs, nothing writes.
One conversation with operations. What you run, what you already trend, and which failures you remember well enough to score us against.
Six to twelve months of BMS trend logs and DCIM history as flat files, under NDA. No connection to your environment, no credential issued, no agent installed.
Models tuned to your plant, your naming conventions and your asset classes, with sensitivity set against your real cost of intervention.
A retrospective against incidents in your own history. Hits, lead times and misses on the same page. No exclusivity, no obligation, and no invoice for any of it.
Forty-five minutes with operations, six to twelve months of history under NDA, and a scored retrospective against incidents you already remember.
No cost to evaluate. No live connection required. No obligation.