Facilities Dive's coverage lays out a problem every data center operator already feels: legacy facilities software was built for slow-moving office buildings. Data centers behave differently. They generate revenue instead of sitting on the books as a pure cost center. The piece, written by Planon partnerships director Bob Mostachetti, cites JLL's 2026 Global Data Center Outlook. JLL puts the industry on pace for 14% compound annual growth through 2030. Nearly 100 gigawatts of new capacity and $3 trillion in investment are landing in that window.
Mostachetti's fix targets a specific failure inside that software: alarms that need a human to read, interpret, and turn into a work order. He wants facilities systems wired directly into DCIM, building management systems, and electric power management systems, so an alarm routes its own work order without a person translating it. He points to Johnson Controls' Metasys 16.0 update as one vendor already chasing that integration, aimed at cutting downtime and easing compliance. He also flags multi-tenant colocation, where SLA reporting and access controls have to respect tenant boundaries that most general-purpose platforms were never built to track.
Mostachetti doesn't name cooling specifically, but his argument lands hardest there. A CDU throwing a flow alarm, a leak sensor tripping on a manifold, or a coolant chemistry reading drifting out of spec all produce the same failure he describes: an alarm that needs a human translator before a technician gets dispatched. Direct-to-chip and immersion deployments multiply that alarm count per rack far past anything a building full of air handlers and rooftop units ever produced. The flow balancing problem already documented in direct-to-chip systems shows how thin the margin for manual monitoring has gotten at rack level.
Coolant chemistry monitoring and flow measurement already generate their own data streams, separate from the BMS and DCIM systems Mostachetti wants unified. Spectrometer-based coolant chemistry monitoring and zero-diameter flow measurement both produce readings that need to land somewhere a facilities team can act on fast. If that data sits in a standalone dashboard instead of a shared work order system, the alarm turns into noise instead of a dispatched technician.
A single colocation hall running liquid cooling for one tenant and air cooling for another needs software that can tell a CDU alarm from an air handler alarm and route each to the right vendor contract. Operators locking in CDU vendors and monitoring platforms this year are making that choice without knowing whether the facilities software underneath can ingest what those systems produce. Closing that gap costs far less before the concrete sets than after.