Your BAS alarms when a value crosses a threshold. But the most expensive problems never cross one — they run inside the normal band, for years. Prescriptiv, our fault-detection engine, reads the trend data your BAS already produces and surfaces them, ranked by what they cost.
Prescriptiv screens your trend data for each of these patterns. The library grew out of Carnegie Mellon research and keeps expanding as deployments surface new failure modes.
Two systems fighting in the same air — comfort holds, so nobody sees it, but you pay to heat and cool at once.
Dampers stuck or mis-sequenced, paying for mechanical cooling during hours the weather would cool for free.
Equipment running nights, weekends, and early mornings for spaces that are empty.
Sensors reading wrong, so controls act wrong everywhere downstream — while the BAS reports normal.
Overrides and one-off tweaks that never got reset, compounding quietly year over year.
Cycling, short-starts, and hours far beyond what the load actually requires.
Loops that hunt, resets that never engage, sequences that drifted from their design intent.
Flow stations reporting fiction — boxes starve or overblow while the BAS shows everything normal.
You don't get a wall of alerts. You get a prioritized list — each finding with the equipment, the evidence, the recommended fix, and the annual dollar value.
Sample finding set · ranked by annual impact
| # | Finding | Equipment | Fix effort | Est. annual impact |
|---|---|---|---|---|
| 01 | Simultaneous heating and cooling, weekday occupied hours | RTU-2 | Sequence change | $18,400 |
| 02 | Economizer at minimum position through free-cooling weather | AHU-1 | Repair / recommission | $11,900 |
| 03 | Occupied schedule running 4am and weekends, space unoccupied | VAV-2-08 | Schedule change | $6,200 |
| 04 | Discharge-air reset disabled after a sequence change | AHU-2 | Re-enable reset | $4,100 |
| 05 | Chilled-water pumps staging early against false demand | CHW plant | Staging tune | $3,800 |
Representative of real deployments. Most top findings are schedule and sequence changes a controls contractor can make in normal maintenance windows — not capital projects.
Developed from Carnegie Mellon research, Prescriptiv evaluates every relevant point against the fault library — relationships and patterns a threshold alarm structurally can't see. See how it fits the platform
Every finding is checked by LeanFM engineers who do commissioning — so what reaches you is real, prioritized, and actionable, not a noisy feed of false positives.
$0
identified in a single 220,000 sq ft school — led by exactly these faults, none of which tripped an alarm.
Read the K-12 findingsSend us one building's trend data and we'll return the findings — ranked, in dollars, with the evidence. No new hardware.