A chief engineer at a large research university in Ann Arbor had a maintenance backlog of 340 open work orders and a deferred maintenance budget that covered maybe 60% of what needed to be done this fiscal year. He had to decide which 40% of needed work to defer — and the honest answer was that nobody on his team had a systematic way to rank the risk of deferring a chiller overhaul vs a boiler heat exchanger inspection vs an elevator rope replacement. Every piece of equipment had an advocate, every department had a priority request, and every decision felt like a judgment call made with incomplete information. That is the problem that AI failure mode analysis solves for campus facilities — not prioritising by complaint volume or department influence, but by calculating the actual probability that deferring a specific maintenance item will result in equipment failure, and ranking the entire backlog by failure risk so budget decisions have a defensible analytical foundation. OxMaint AI failure mode analysis ranks every deferred maintenance item and open work order by collision risk, failure probability, and consequence severity — giving facilities directors the data to make defensible priority decisions instead of judgment calls under budget pressure.
AI Ranks Every Deferred Maintenance Item by Failure Risk — So You Know What to Fix First
OxMaint AI FMEA scores every campus equipment item by failure probability, consequence severity, and detectability — turning 340 open work orders into a ranked priority list with a defensible analytical basis
340
Open work orders — Ann Arbor research university, 60% budget coverage — which 40% to defer with no systematic risk ranking
3× RPN
OxMaint AI FMEA Risk Priority Number — combines failure probability, consequence severity, and detectability into one rankable score per equipment item
62%
Of deferred maintenance failures at campus facilities were preventable if the highest-RPN items had been prioritised ahead of lower-risk deferred work
Six Campus Equipment Categories Where AI FMEA Changes Maintenance Decisions
FMEA (Failure Mode and Effects Analysis) has been an engineering tool for decades — used in aerospace, automotive, and defence to rank failure risks systematically. AI-powered FMEA in OxMaint applies the same methodology to campus critical equipment using sensor data, maintenance history, and equipment age to calculate current RPN scores that reflect actual condition, not theoretical failure rates. OxMaint AI FMEA runs continuously on all six critical equipment categories — updating RPN scores as conditions change.
Central Chillers & Cooling Towers
Highest consequence — campus-wide cooling failure
Chiller failure during summer semester at a research university affects laboratory environments, data centers, and occupied academic buildings simultaneously. AI FMEA in OxMaint tracks compressor vibration trends, refrigerant pressure history, oil analysis results, and bearing temperature data to calculate a current RPN score that reflects the chiller's actual failure probability — not just its age or PM interval remaining.
Steam Boilers & Heat Exchangers
Safety-critical — pressure vessel failure consequence
Steam boiler failure consequences range from building heating loss to pressure vessel incidents with life safety implications. OxMaint AI FMEA tracks combustion efficiency trends, pressure relief valve test history, heat exchanger water quality results, and tube inspection findings to calculate RPN scores that separate boilers approaching failure from those with genuine remaining safe operating life — preventing both under- and over-maintenance decisions.
Elevators & Vertical Transportation
Compliance + safety — rope condition and brake tests
Elevator failure consequences include ADA compliance violations, occupant safety events, and state inspection failures that take an elevator out of service for weeks. OxMaint AI FMEA tracks rope wear measurements, brake adjustment history, door mechanism test results, and controller fault log patterns to rank elevator maintenance priority by actual safety risk — identifying equipment approaching regulatory non-compliance before the state inspector does.
Emergency Power & Generators
Life safety — failure during grid outage
Emergency generator failure during a power outage affects life safety systems, medical facilities, and laboratory equipment with refrigerated samples simultaneously. OxMaint AI FMEA analyses load test results, fuel quality data, coolant condition, and battery test outcomes to rank generator sets by actual readiness — not by test schedule compliance alone. A generator that passes monthly runtime tests but has degrading coolant and marginal battery voltage is flagged as high RPN despite apparent compliance.
Air Handling Units & VAV Systems
Air quality and occupant comfort — IAQ consequences
AHU failures cause indoor air quality problems in occupied academic spaces that result in classroom closures, occupant health complaints, and HVAC system balancing failures that affect entire buildings. OxMaint AI FMEA tracks belt wear patterns, coil fouling indicators, damper actuator test results, and filter pressure differential trends to identify AHUs approaching failure — particularly those serving laboratories with fume hood make-up air requirements where failure consequences exceed simple comfort issues.
Electrical Switchgear & Transformers
Power distribution — building and campus zone outages
Campus electrical switchgear failures cause building-level or zone-level power outages that affect teaching, research, and residential operations simultaneously. OxMaint AI FMEA tracks partial discharge trending, insulation resistance test history, thermographic survey results, and oil analysis data to rank electrical equipment by actual failure proximity — enabling targeted maintenance investment on equipment genuinely approaching failure rather than uniform time-based maintenance across all panels.
OxMaint — AI FMEA for Campus Facilities
Rank 340 Work Orders by Failure Risk in Minutes — Not Weeks of Engineering Analysis.
Chillers, boilers, elevators, generators, AHUs, and electrical — AI RPN scores for every critical asset, updated continuously from sensor and maintenance data.
Traditional FMEA vs OxMaint AI FMEA — Speed and Accuracy
Traditional FMEA is a manual engineering exercise — workshops, expert interviews, and static spreadsheets that are accurate on the day they are produced and outdated the day after. OxMaint AI FMEA updates continuously from live sensor data and maintenance records. OxMaint AI FMEA gives facilities directors a live risk ranking that reflects current equipment condition — not a six-month-old workshop output.
Traditional FMEA
Manual workshop — engineering team, static output
Time to complete FMEA for 200-asset campus
6–12 weeks
RPN score accuracy after 90 days
Degrading — conditions change
Cost to produce initial FMEA
$28,000–$80,000
Update frequency
Annual or less
Maintenance History Only
CMMS work order patterns — no sensor data
Time to complete ranking for 200-asset campus
1–2 weeks manual analysis
RPN score accuracy
Moderate — lagging indicator only
Cost to produce ranking
$4,000–$12,000 staff time
Update frequency
Quarterly with effort
OxMaint AI FMEA
Live sensor + maintenance history — continuous RPN
Time to initial ranking for 200-asset campus
Under 48 hours after data connection
RPN score accuracy
Current — updated with every sensor read
Cost to produce and maintain ranking
Included in OxMaint platform
Update frequency
Continuous — live dashboard
Technology Stack: How OxMaint Calculates AI FMEA RPN Scores
OxMaint AI FMEA RPN is calculated from four connected data sources that, together, provide a more accurate failure risk picture than any single data type could alone. The RPN score is a composite of occurrence probability (O), severity of failure consequence (S), and detectability of early warning signals (D) — calculated from live data, not engineering assumptions. Connect your campus systems to OxMaint AI FMEA.
IoT Sensor & PLC Data — Occurrence (O) Score
Vibration sensors, temperature transmitters, pressure transducers, and PLC fault log streams feed OxMaint continuously. AI analyses trend patterns — bearing temperature rate of increase, vibration amplitude trend slope, pressure differential drift — to calculate the Occurrence probability score (1–10) that reflects how likely a failure is based on current operating condition trends rather than historical fleet failure rates.
AI Camera Vision — Detectability (D) Score
OxMaint AI cameras at critical equipment locations detect visible failure precursors — oil leaks, insulation damage, corrosion, missing guards, and discoloration — that inform the Detectability score (1–10). Equipment where AI camera vision provides early warning of developing faults scores lower on Detectability (good) than equipment where failure mode provides no visible warning before functional failure occurs.
Asset Criticality Model — Severity (S) Score
OxMaint's asset criticality model assigns Severity scores (1–10) based on the equipment's role in campus operations — life safety systems, research laboratory support, residential HVAC, and general academic occupancy are scored differently because their failure consequences are different. Severity scores are configured per building type and can be adjusted by the facilities director to reflect the institution's specific risk priorities and operational context.
SAP & ERP Integration — Budget vs RPN Optimisation
OxMaint AI FMEA RPN scores sync with SAP Plant Maintenance and capital planning systems — enabling budget decisions that are directly informed by risk ranking. Deferred maintenance items are ranked in SAP by their OxMaint RPN score, so the highest-risk items surface automatically to the top of the capital request list. Facilities directors present budget requests with AI risk evidence, not advocacy arguments.
"Every budget cycle, every department argued their equipment was the priority. With OxMaint AI FMEA, I show the VP a ranked list with an RPN score for every item. The chiller with a 720 RPN gets funded before the AHU with a 210 RPN — no matter how loudly the building manager argues. We stopped having the same arguments every year."— Chief Facilities Engineer, Research University · 48 buildings · Michigan, USA
Frequently Asked Questions
Q1How does OxMaint AI FMEA calculate the RPN score — and what data does it require?▼
OxMaint RPN = Occurrence (O) × Severity (S) × Detectability (D), each scored 1–10. O is calculated from sensor trend data and maintenance history. S is configured from the asset criticality model per equipment type and building role. D is assessed from available monitoring coverage and early warning signal quality. Minimum data requirement is maintenance history — sensor data improves accuracy significantly but is not required for initial deployment.
Q2Can OxMaint AI FMEA differentiate between failure modes on the same piece of equipment?▼
Yes — OxMaint supports multiple failure mode records per asset, each with its own O, S, and D scores and associated sensor or maintenance triggers. A chiller may have separate RPN scores for compressor bearing failure, refrigerant leak, and motor winding insulation failure — each tracked independently and each generating separate maintenance recommendations when RPN thresholds are crossed.
Q3How does OxMaint FMEA handle equipment with no sensor monitoring — maintenance history only?▼
For assets without sensor monitoring, OxMaint AI calculates Occurrence probability from maintenance history patterns — failure frequency, repair interval trends, and parts consumption velocity. The RPN score is less precise than with sensor data but still provides a meaningful relative ranking across the asset portfolio. As sensors are added, the RPN model automatically incorporates the new data stream.
Q4Can facilities directors adjust the Severity scoring to reflect their specific campus risk priorities?▼
Yes — Severity scores in OxMaint are configurable per asset class and building type. A facilities director can assign higher Severity weights to research laboratory HVAC than to administrative building systems, reflecting the institution's actual consequence priorities. Custom Severity configurations are stored per asset and applied automatically to all future RPN calculations without requiring manual override.
Q5How does OxMaint FMEA integrate with the capital planning and budget request process?▼
OxMaint generates a deferred maintenance risk report ranked by RPN score — exportable as a PDF or Excel file with asset description, current RPN, failure mode, estimated repair cost, and consequence description for each item. This report format is directly usable as supporting documentation for capital budget requests to university administration, providing the risk-evidence basis that replaces informal advocacy in the budget allocation process.
OxMaint — AI FMEA Campus
Stop Guessing Which Equipment to Fix First. Let AI Show You.
48 hrs
to initial RPN ranking
Live
continuous updates
Free
to start today
