The organization
A colocation data center rents racks, cages and power to enterprise customers. Its facilities team keeps the power chain and cooling running around the clock.
Customers sign contracts that promise power and temperature within agreed limits, and they ask for maintenance records during their own audits.
The challenge
Critical equipment is serviced by vendor visit dates, generator tests are logged on paper, and monitoring alarms land in an inbox with no record of the response.
- UPS and CRAC services follow the calendar, not the hours the units have run.
- Monthly generator tests are written on paper sheets that are hard to produce for an audit.
- Temperature and humidity alarms are emailed, and nobody owns the follow-up.
- Vendor work and in-house work are tracked in different places.
- Customers ask for maintenance evidence that takes days to compile.
How it is set up in Asset Infinity
The data center runs its critical facilities in Asset Infinity, with each UPS, generator and CRAC unit in one register.
Runtime-based plans
Generators and CRAC units have plans that fall due on runtime hours, alongside date-based plans for UPS batteries and annual vendor services. (Maintenance)
Generator tests as inspections
The monthly load test is a checklist with required readings, photos and a signature. A failed check raises corrective work automatically. (Maintenance)
Sensor limits in the halls
Temperature and humidity sensors have limits. A reading outside them raises a work order for the CRAC unit serving that zone. (Connected)
Records ready for audits
Dashboards show plans due and open alarms, and reports pull each asset’s tests and repairs for a customer audit. (Insight)
How it works day to day
Here is how a month of critical facilities work runs in Asset Infinity.
- System: watches the halls. Temperature and humidity sensors report from each zone. A reading above its limit raises a work order for the CRAC unit serving that zone.
- Shift engineer: responds to the alarm. The engineer is notified on the mobile app, checks the unit, clears a blocked filter, records the readings and closes the work order.
- Shift engineer: runs the generator test. On the monthly date, each generator’s test inspection falls due. The engineer records load, voltage, frequency, fuel level and the hour meter, then signs.
- System: raises corrective work. A failed check, such as a battery charger fault, raises a corrective work order for that generator.
- System: counts runtime. The hour meter readings feed runtime plans. When a generator or CRAC unit reaches its interval, a preventive work order is created.
- Facilities manager: books the vendor. Work needing the generator vendor waits on the supplier, with its SLA clock paused, and reminders go out if it sits too long.
- Account manager: answers an audit. A report lists every test, alarm and repair for the equipment serving a customer’s cage, with readings and signatures.
Before and after
| Before | With Asset Infinity | |
|---|---|---|
| Service timing | Vendor visit dates | Runtime hours and dates per unit |
| Generator tests | Paper sheets in a binder | A signed inspection with required readings |
| Failed test | Noted, then followed up by memory | Raises a corrective work order |
| Environmental alarms | Emails nobody owns | A work order with an owner |
| Customer audits | Days compiling records | One report per asset or cage |
The first month
- Critical assets loaded. UPS, generators, CRAC units, PDUs and switchgear are imported by hall, with labels on each unit.
- Plans and checklists set. Runtime and date plans are created, and the monthly generator test becomes an inspection checklist.
- Sensors connected. Hall temperature and humidity sensors are linked to their zones, with limits set.
Example results
- Every generator tested monthly, with readings and a signature on record
- Every temperature or humidity alarm raises a work order within minutes
- 100% of UPS, generator and CRAC units on a preventive plan
- Each alarm has an owner, a response time and a record of what was done.
- Service timing follows the hours each unit has actually run.
- Jobs waiting on a vendor pause their SLA clock and show who they wait on.
- Audit evidence for a customer comes from one report instead of a file hunt.
Why it worked
Nothing waits on memory
Alarms, failed checks and due plans each create a work order, so the response does not depend on who is on shift.
Service by use
Runtime-based plans service generators and CRAC units when their hours say so, not just when the calendar does.
Evidence on demand
Readings, photos and signatures are stored with each asset, so customer audits take a report instead of a week.
Looking ahead
With power and cooling equipment in one register, the facilities team knows what is due, what is alarming and what has been tested, at any hour.
The same records answer customer audits, and the repair history of each unit shows which equipment is reaching the end of its life and belongs in next year’s capex plan.