How The Xavier University Physical Plant Turned a 2-Hour Search Into a 10-Second Answer
Prepared by: Jacob Homan, Founder · Operis Facility Technologies
Subject institution: Xavier University Physical Plant, Cincinnati, Ohio
Pilot period: 90 days · 54 buildings · 2.5 million square feet
Xavier University Physical Plant operates a 54-building, 2.5 million square foot campus with two central utility plants. In a 90-day pilot, the department deployed FRED, an AI knowledge assistant developed by Operis Facility Technologies, to give field technicians immediate access to the building documentation they depend on daily.
Within 90 days, 24 team members were actively using the system, the document library grew from 9 files to over 400, and questions that previously required 1 to 3 hours to resolve were being answered in under 10 seconds. Technicians stopped driving back to the shop to find documents. New hires began using the system for campus orientation without interrupting experienced staff.
This document summarizes the context, approach, key findings, and implications for other facilities teams considering a similar deployment.
Key Findings
Background
Xavier University Physical Plant manages 54 buildings across a 2.5 million square foot campus, with two central utility plants and a facilities team responsible for everything from daily maintenance to emergency response. Like most university physical plants, the department had accumulated years of documentation: O&M manuals, HVAC as-builts, floor plans, equipment schedules, SOPs, and spreadsheets containing critical building data.
The problem was not that the information did not exist. It was that accessing it from the field required time most technicians did not have. A technician with a question had three options: drive back to the shop and search through files, call a supervisor or coworker and hope they knew the answer, or make a judgment call without the information they needed. None of these options were fast, and none of them scaled as the team grew.
Implementation Approach
The pilot began with 9 documents uploaded to a private FRED account on day one. No reformatting was required. No IT involvement was needed. Documents were uploaded as-is: PDFs, Excel spreadsheets, CAD-derived floor plans, and Word files.
Team members accessed FRED from their phones in the field, asking questions in plain English. FRED returned answers cited directly from the source documents, including the file name and page number. Adoption was voluntary. Orientation was brief. Usage grew because the tool worked.
Field Questions and Outcomes
The questions asked during the pilot illustrate the operational value more precisely than aggregate metrics. The following exhibit presents selected real questions asked by Xavier Physical Plant staff during the 90-day period, alongside the time the same retrieval would have required using prior methods.
| Question Asked | Response Time |
|---|---|
| "Provide the HVAC drawing for the 2nd floor of the Cintas Center." | Under 10 sec Previously: 1–2 hours |
| "What is the capacity of the elevator at McDonald?" | Under 10 sec Previously: 30–60 min |
| "Where is the circuit located for Parking Lot A?" | Under 10 sec Previously: 1–3 hours |
| "Where is the water shutoff for the dorm rooms on the 3rd floor?" | Under 10 sec Previously: 15–30 min |
| "What is the contact for the chiller service contractor?" | Under 10 sec Previously: 10–30 min |
| "Provide the map to Cohen Center." / "What is the address to Elet Hall?" | Under 10 sec Previously: interruption + 5–15 min |
Knowledge Base Growth and Organic Adoption
By the end of the pilot, the Xavier FRED knowledge base had grown from 9 files to over 400 documents, drawings, and manuals. This growth was not directed by management. It was driven by the team itself.
The clearest evidence of organic adoption was a behavior that emerged without prompting. Superintendents and technicians began submitting information to the knowledge base with the subject line "Food for FRED." When they encountered a question the system could not yet answer, or came across information the team should have access to, they flagged it for inclusion. They were not simply using the system. They were investing in it.
Technology adoption in facilities management is often slow and skeptical, particularly among field technicians who have seen software initiatives come and go without improving their day-to-day work. The Food for FRED pattern indicates the system had crossed a meaningful threshold: the team believed in it enough to spend their own time making it better.
| Document Category | Contents |
|---|---|
| General floor plans | 200 documents covering the full campus building inventory |
| HVAC as-built drawings | 200+ mechanical drawings across all major buildings |
| MEP manuals | Multiple manuals including a 400+ page manual for an on-campus aquatic facility |
| Equipment and location spreadsheets | 25+ Excel files covering water shutoffs, gas shutoffs, panel locations, contractor contacts, building addresses, and key personnel |
| Equipment specifications | Full specs for all elevators, backflow preventers, and filter sizes across the campus |
| Procedures and policies | Elevator entrapment procedures, emergency on-call policies, and operational SOPs |
Operational Impact
The change that mattered most was not measured in queries per week. It was measured in where work happened.
Before FRED, a technician with a question in the field had to leave the job to get the answer. They called supervisors. They drove back to the shop. In ceiling-space work, tracking down a valve location might mean climbing a ladder, opening tiles, finding nothing, coming back down, making a call, and climbing again. The answer always required interrupting the work.
With FRED, the answer is in their pocket before they climb the ladder the first time.
For new technicians, the impact was different but equally significant. Campus orientation, building locations, vendor contacts, all the institutional knowledge that new hires typically absorb slowly through osmosis or by interrupting experienced colleagues, became immediately accessible. New team members were retrieving information in their first week that would previously have taken months to accumulate.
FRED should be here.Xavier University Physical Plant technician, during an engineering coordination meeting — meaning the information being discussed belonged in the knowledge base
A technician attending a planning meeting, watching information be discussed and decisions be made, turned to the room and said FRED should be present. He was not talking about the software. He was saying that the knowledge being exchanged belonged in the system so the whole team could access it. The team had stopped thinking of FRED as a search tool. They had started thinking of it as a resource that belongs at the table.
Next Phase: Field Contribution and New Construction Integration
The pilot identified one friction point worth addressing directly. When a technician discovers something in the field, a valve location that was never documented, an updated specification, a note that would help the next person on that job, there was no fast way to get that information into FRED from the field.
The next version of FRED addresses this through a field submission feature. Technicians will be able to submit a photo, voice note, or scanned document from any device. Submissions route to an administrator for review. Once approved, the information becomes a permanent part of the team's knowledge base, built from the field, by the people doing the work.
Xavier is also undertaking a new College of Medicine building project. Operis Facility Technologies is partnering with the general contractor to deliver all O&M documents and as-built drawings directly into FRED at project completion, rather than the conventional handoff of a hard drive containing files that few team members can navigate efficiently. When the building opens, the facilities team will have full document access from day one, before they have set foot in the mechanical room.
Implications for Other Facilities Teams
The conditions at Xavier at the start of this pilot were not unusual. Accumulated documentation spread across binders, shared drives, and institutional memory. Technicians spending meaningful time each day retrieving information rather than doing work. Institutional knowledge concentrated in experienced staff approaching retirement. These conditions describe the majority of university physical plants and facilities operations at comparable scale.
What the Xavier pilot demonstrated is that the adoption barrier is lower than most operations teams anticipate. A complete document library is not required to begin. What is required is the set of documents the team reaches for most frequently, and a system that makes them accessible from any device, from anywhere on campus.
The library grows organically once the team sees value. Xavier's knowledge base grew more than 4,000 percent in 90 days, not because it was mandated, but because the team was invested in it. That investment is the outcome that matters most, because it is the one that sustains the system after the pilot ends.
See how FRED works for your team
Watch the 4-minute demo and see how fast your team could go from binders to answers in the field.
