How AI Is Reshaping Service Operations in Mission Critical Infrastructure

This interview evaluation is sponsored by Aquant and was written, edited, and printed in alignment with our Emerj sponsored content guidelines. Learn extra about our thought management and content material creation companies on our Emerj Media Services page.

Service organizations supporting power, infrastructure, and data-center property face a structural mismatch: uptime necessities are being tightened to near-zero, whereas upkeep fashions and technician capability haven’t saved tempo.

Demand is accelerating quickest on the grid edge. The U.S. Department of Energy cites Electric Power Research Institute evaluation exhibiting information facilities might eat as much as 9 p.c of U.S. electrical energy era by 2030, greater than double their 2023 share. That strain already reveals up in downtime prices: the Institute for Supply Management reported unscheduled downtime now prices the world’s 500 largest firms $1.4 trillion yearly — 11 p.c of income.

Meanwhile, the workforce wanted to forestall that downtime is shrinking. The U.S. Bureau of Labor Statistics projects 81,000 electrician openings yearly by means of 2034, pushed largely by retirements moderately than new entrants.

Fragmented gear information compounds the hole. The National Institute of Standards and Technology found insufficient interoperability of facility and gear information prices U.S. capital-facilities house owners and operators $10.6 billion yearly throughout operations and upkeep alone — the identical fragmentation that leaves technicians with out asset historical past or manuals on the level of service, unable to ship a first-time repair with out next-best upkeep steering.

Without next-best upkeep steering, technicians can not shut this hole alone.

Emerj’s Yolandi de Weerdt was joined by Joe Lang, Vice President of Service Technology and Innovation at Comfort Systems USA, on the AI in Business podcast to delve into how AI is reshaping service operations in mission‑essential infrastructure.

This article examines the operational and strategic insights rising from Joe Lang’s perspective on AI‑enabled service transformation:

• Anomaly detection for situation‑primarily based gear upkeep: Capture actual‑time sensor information so service groups can intervene primarily based on gear habits moderately than scheduled duties, supporting close to‑zero‑downtime environments.
• Prescriptive steering for constant technician efficiency – Consolidate diagnostic proof into real-time, next-best-action suggestions so technicians transfer from forecasting failures to fixing them sooner.
• Operational transformation required for upkeep workflow change: Treat the workflow change as a devoted operational initiative with clear possession, dedicated resourcing, and steady refinement so the group achieves decrease downtime and decrease prices.

Listen to the complete episode beneath:

Episode: How AI Is Reshaping Service Operations in Mission Critical Infrastructure – with Joe Lang of Comfort Systems USA

​Guest: Joe Lang, Vice President, Service Technology and Innovation at Comfort Systems USA

Expertise: Service Technology, AI-Enabled Service Operations, Field Service Innovation, Customer Experience

Brief Recognition: Joe Lang is Vice President of Service Technology and Innovation at Comfort Systems USA, the place he leads know-how and innovation initiatives centered on advancing subject service operations and buyer expertise. He has spent greater than 18 years with Comfort Systems USA in govt service management roles and beforehand held management positions at Johnson Controls and York International, overseeing service operations and enterprise progress. Lang additionally serves on the advisory boards of Field Service USA, The Service Council, and Aquant. He holds a bachelor’s diploma in Industrial Technology from Purdue University.

Anomaly Detection for Condition‑Based Equipment Maintenance

Maintenance cycles assume predictable intervals, however gear habits usually modifications in the unmonitored durations between them. When these anomalies are left unaddressed, they progress into gear failure and avoidable downtime. That hole between deliberate upkeep and actual‑time actuality is the operational threat Joe Lang zeroes in on and the explanation he argues that organizations should deal with actual‑time habits as their supply of reality.

His distinction in the episode is clearly that anomaly detection isn’t a complicated AI function, however the first operational self-discipline that lets groups act on rising points earlier than they escalate into failures.

Lang emphasizes that enterprises already accumulate the required sensor information; what’s lacking is the rigor to floor deviations early sufficient for technicians to intervene. He argues that the majority failures aren’t surprises — they’re detectable behavioral drifts that organizations merely aren’t appearing on. For him, anomaly detection is the second service work turns into proactive moderately than reactive.

Lang frames the operational stakes:

AI provides technicians a head begin. When the system flags a deviation, it’s usually the earliest signal that one thing is drifting out of regular habits. Acting at that second prevents failure moderately than reacting to it. It modifications the rhythm of service work — groups cease chasing emergencies and begin addressing points earlier than they turn out to be essential.

— Joe Lang, Vice President of Service Technology and Innovation, Comfort Systems USA

Joe identifies operational necessities for adopting anomaly detection:

• Instrument precedence property: Deploy and validate sensors on gear the place downtime creates the very best operational or contractual threat.
• Define deviation thresholds: Establish clear behavioral triggers — temperature, vibration, strain, load — that sign actionable drift moderately than noise.
• Automate technician routing: Ensure deviations generate service duties instantly, with out handbook overview or batching.
• Measure intervention timing: Track how shortly groups reply to deviations and correlate early interventions with fewer failures averted and fewer emergency dispatches.

These capabilities set up the working ground Lang argues for, a service group that responds to actual‑time habits moderately than scheduled assumptions.

Prescriptive Guidance for Consistent Technician Performance

Lang attracts a pointy distinction between predictive and prescriptive upkeep — the shift from forecasting failures to recommending the next-best motion. He notes that many organizations nonetheless exchange parts like air filters on mounted schedules even when pressure-drop information reveals they’re working clear, a spot prescriptive steering closes by aligning interventions with precise gear habits moderately than calendar assumptions.

Technician efficiency swings after they face unfamiliar gear, ambiguous signs, or failure modes that current every time in a different way. Lang’s level is that this variability isn’t a personnel difficulty, however an info difficulty. Technicians begin from totally different baselines as a result of the diagnostic proof they want is scattered throughout disconnected programs, buried in PDFs, or locked inside particular person expertise.

Prescriptive steering stabilizes that variability by giving each technician the identical knowledgeable place to begin. When service histories, OEM documentation, decision patterns, and gear context are consolidated and delivered in actual time, the system can floor not simply the doubtless fault however the next-best motion to resolve it — earlier than the panel is opened. Technicians nonetheless make the decision, however they start with the group’s accrued diagnostic intelligence moderately than guesswork.

Lang emphasizes that this isn’t about changing technician judgment. It’s about eradicating the primary ten minutes of uncertainty that drive inconsistent outcomes. When the doubtless fault and really helpful intervention arrive in the meanwhile of service, diagnostic swings slim, first-time-fix outcomes rise, and a constrained workforce operates with a steadier baseline of efficiency.

Evidence sources Joe identifies for prescriptive steering:

• Service histories: Real failure modes, signs, and corrective actions that present how points really offered and the way they had been resolved.
• Manufacturer documentation: OEM steering that defines meant habits, recognized fault pathways, and validated diagnostic steps.
• Resolution patterns: Proven fixes that constantly resolved points in the sphere, revealing interventions with a dependable monitor document.
• Equipment context: Identity, configuration, and working situations that guarantee suggestions mirror the asset’s precise habits.

To make these proof sources usable in the meanwhile of service, Lang stresses they need to dwell in a single structured information surroundings that the mannequin can purpose over and ship again to technicians in actual time.

Joe identifies operational data-workflow design selections for making prescriptive steering efficient:

• Consolidate diagnostic proof: Bring service histories, OEM documentation, and determination patterns into one structured surroundings so the mannequin can purpose throughout the complete diagnostic proof set.
• Deliver steering in actual time: Surface the doubtless fault and the next-best motion immediately in the technician’s workflow — the cellular app, work order, or dispatch interface.
• Anchor suggestions to the gear context: Ensure steering displays the precise unit’s identification, configuration, and behavioral historical past moderately than counting on generic assumptions.

These diagnostic-workflow design selections create the constant place to begin Lang argues technicians want — a real-time, evidence-driven baseline that reduces variability and strengthens first-time-fix efficiency throughout a constrained workforce.

Operational Transformation Required for Maintenance Workflow Change

Lang emphasizes that modernization begins with figuring out, categorizing, and organizing property into logical gear teams. Without structured asset information, centralized manuals, and repair historical past, organizations can not reliably apply anomaly detection or prescriptive steering.

Lang describes that this effort sometimes breaks down since organizations are likely to deal with the transition as a part-time initiative — pulling a batch of information right here, loading it right into a platform there — moderately than resourcing it as devoted work with clear possession.

He argues this under-resourcing is the most typical purpose modernization efforts stall: groups spend a yr assembling information infrastructure however produce no measurable outcome, as a result of the mission was by no means staffed to succeed.

Lang frames the operational actuality:

This is the place you’ll modify the aircraft when you’re flying it. You’ve acquired to switch it so it may proceed to fly and land and take off once more. You completely should useful resource this appropriately if you begin down this path.

— Joe Lang, Vice President of Service Technology and Innovation, Comfort Systems USA

Joe identifies operational necessities for treating upkeep workflow change as a devoted initiative:

• Assign devoted possession: Staff a selected staff or mission lead accountable for the transition, moderately than distributing the work throughout present roles as a secondary job.
• Inventory and categorize property first: Group gear by kind and element similarity earlier than making an attempt to use anomaly detection or prescriptive fashions throughout the fleet.
• Centralize manuals and repair historical past: Assemble OEM documentation and gear data right into a database technicians can entry immediately in the sphere, moderately than looking for info after arriving on-site.
• Commit enough resourcing: Treat the initiative as a completely staffed program, not an incremental information mission layered onto present workloads.

For Lang, the mix of dedicated resourcing and structured asset information determines whether or not a company really reduces downtime and prices or just accumulates unused information infrastructure.