# In FMJ, a hypothetical cancelled class triggers nine actions By MarketScale Newsroom · Published 2026-10-01 · Building Management on MarketScale Canonical: https://www.marketscale.com/industries/building-management/in-fmjs-ai-building-one-cancelled-class-sets-off-nine-coordinated-actions > David Rowden’s FMJ feature uses a hypothetical: one cancelled class and a nine-action chain across a campus. It’s a practical vendor test. ## Key points - One occupancy reading in an empty lecture hall can reach cleaning crews, security patrols, parking and the energy forecast, so an orchestration platform likely needs sign-off from more than the facilities team. - A useful AI maintenance alert can give a probability and time window—e.g., a 78% chance of failure inside 28 days—so it beats a raw trend line that someone must notice. - The sharper vendor question is which of the nine actions in the cancelled-class scenario a platform can carry out on its own today, which it can only recommend, and which systems it writes to rather than just reads. Rowden's broad claim will sound familiar to anyone who has sat through a smart-building pitch. The lecture-hall scenario is the useful part, because it turns a vague promise into something a facilities director can hold a vendor to. Rowden also notes that organizations around the world face growing pressure on several fronts at once: cutting operating costs, raising energy performance, showing ESG outcomes, becoming more resilient, improving occupant well-being and extending asset life. At the same time, the building generates a constant stream of data from BMS, CMMS, CAFM and IWMS platforms, IoT devices, CCTV analytics, access control and environmental sensors. There is no shortage of data. The hard part is getting the systems to act on it together. ## Systems that frequently operate in isolation People absorb the cost of that isolation. FM teams end up reading several dashboards, sorting out alarms that contradict each other and coordinating the response by hand. He draws the line against automation with a deliberately plain rule: if a room goes above 24 C, turn on cooling. One condition, one response. Orchestration starts from a single signal and coordinates many responses across systems that were never designed to coordinate. Rowden writes that automation has delivered operational improvements. His argument is that each rule stops at the edge of its own system. ## What the empty lecture hall touches It's more revealing to sort that list by who owns each action than by which system runs it. Only a few of the items sit squarely with the people who run the chillers. On a campus where cleaning, security and parking are separate operations, one occupancy reading now reaches all three, plus whoever owns the energy forecast and the booking calendar. Orchestration is an org-chart decision as well as a software purchase. If a security manager hasn't agreed to let software change patrol routes, the platform can only recommend that change, not make it. There's a fair objection. Rowden presents the scenario as a hypothetical, and the feature names no platform, savings range or integration method, so the nine-action chain is a claimed capability, not a measured result. For a building whose main problem is a handful of badly tuned rules, plain automation may be enough. The case for orchestration is strongest where many systems have to respond to the same event, as on a campus with a busy timetable. ## Alerts that arrive already ranked Rowden's second change is to what reaches the facility manager. Instead of thousands of raw alerts, AI would hand over prioritized recommendations backed by predictive analysis. Block Field A maintenance planner can schedule against a 78% chance of failure inside 28 days. A vibration trend on a dashboard still needs someone to read it first. One of Rowden's examples is a current maintenance backlog that increases failure risk for electrical assets serving critical clinical spaces. That suggests ranking work by what an asset keeps running as well as by its condition. It could capture the kind of judgment a senior engineer makes from memory, written down where a system can use it. Felix Oluwalomola, writing in FMJ, noted that run-to-fail maintenance can look cheaper but may bring higher long-term costs and operational setbacks. ## Running a platform through the lecture-hall test Block Field That question separates a dashboard that sees everything from a layer that can act. It also brings the integration work into view, since the nine actions are spread across booking, parking, security, lighting, HVAC and maintenance systems from Rowden's own inventory. The clinical example shows who would be ready first. Ranking a backlog by the spaces each asset serves would depend on the CMMS, which Oluwalomola describes as software that streamlines and centralizes maintenance operations, including work orders, equipment management and planned maintenance scheduling. Hospitals and campuses that already map assets to the spaces they serve are best placed to test Rowden's model. For everyone else, building that mapping comes before choosing an AI platform. ## Sources - [The Intelligent Facility - fmj.ifma.org](https://fmj.ifma.org/the-intelligent-facility-2026) (FMJ Magazine) - [Smart Facilities, Smarter Decisions - FMJ Magazine](https://fmj.ifma.org/smart-facilities-smarter-decisions) (FMJ Magazine) - [IFMA's FMJ Previous Issues](https://www.ifma.org/content-channels/ifmas-fmj/fmj-previous-issues/) Tags: FMJ Magazine, IFMA, facility management, intelligent orchestration, AI in buildings, building management systems, predictive maintenance, CMMS, HVAC controls, smart buildings, facilities operations, technology procurement --- Source: MarketScale, https://www.marketscale.com/industries/building-management/in-fmjs-ai-building-one-cancelled-class-sets-off-nine-coordinated-actions. Published for AI indexing and citation; cite the canonical URL. Site guide for agents: https://www.marketscale.com/llms.txt