IoT-Driven Facilities Management: Cutting Maintenance Costs 38% at a Major Airport
Challenge
Aerolia managed 1.4 million sq ft of terminal space using a reactive maintenance model. Frequent failures in HVAC, escalators, lifts, and baggage systems caused passenger disruption and costly emergency call-outs. Annual unplanned maintenance spend reached £4.2M.
Solution
We deployed an IoT sensor network across 340 critical assets, integrated with a predictive maintenance platform analysing vibration, temperature, and power data to detect failure patterns 2–4 weeks in advance.
Results
Unplanned maintenance incidents reduced by 52%. Annual maintenance costs reduced from £4.2M to £2.6M. Fault response time reduced to 18 minutes. Escalator and lift availability increased from 94.2% to 99.1%.
The Cost of Reactive Maintenance in a 24/7 Environment
Airports do not have downtime windows. When infrastructure fails, the impact is immediate:
- passenger congestion
- missed connections
- operational delays
- reputational damage
At Aerolia:
- 1.4M sq ft of operational terminal space
- 340 critical mechanical and electrical assets
- 23 escalator and lift failures per month
Each failure triggered:
- emergency engineering call-outs (premium cost)
- passenger disruption during peak traffic windows
Annual impact:
£4.2M in unplanned maintenance spend
The mandate: Reduce failures by at least 40% within 18 months
System Design Approach
This was not a monitoring project. It was a shift from:
reactive maintenance → predictive operations
The platform needed to:
- detect failure before it happens
- prioritise intervention by risk
- integrate directly into operational workflows
Sensor Infrastructure
We instrumented 340 assets across five categories, each requiring tailored sensing strategies.
1. HVAC Systems
- vibration sensors (compressors, fans)
- temperature and humidity probes
- power consumption monitoring
Identifies inefficiencies and early mechanical wear
2. Escalators
- motor vibration monitoring
- handrail drive analysis
- chain tension sensors
- step gap safety detection
Critical for passenger flow and safety compliance
3. Passenger Lifts
- motor current signature analysis
- door cycle tracking
- levelling accuracy monitoring
Detects degradation before service disruption
4. Baggage Handling Systems
- vibration monitoring on drive motors
- thermal sensing for overheating detection
Prevents cascading operational failures
5. Electrical Infrastructure
- power quality monitoring
- harmonic distortion detection
- phase imbalance tracking
Identifies upstream electrical issues affecting multiple systems
Deployment approach:
- phased rollout over 14 weeks
- zero operational downtime
- installation aligned with maintenance windows
Predictive Maintenance Model
Each asset type required a custom failure prediction model.
Training Data Sources
- 3 years of historical maintenance logs
- manufacturer failure mode specifications
- operational data (usage cycles, load, environment)
Model Outputs
- Daily asset health score
- Failure probability window (14-day horizon)
- Alert threshold triggers
Maintenance teams receive alerts 2–4 weeks before failure
Operational Integration
Prediction without action has no value.
CMMS Integration (IBM Maximo)
- predictive alerts automatically generate work orders
- prioritisation based on risk and impact
- unified view of:
Enables proactive scheduling instead of emergency response
Performance Outcomes
Failure Reduction
- Unplanned incidents ↓ 52%
Cost Reduction
- Maintenance spend ↓ from £4.2M → £2.6M
- Total reduction: 38%
Operational Efficiency
- Average response time: 18 minutes
Asset Availability
- Escalators & lifts:
Operational Impact
Before:
- Reactive maintenance model
- High emergency call-out costs
- Frequent passenger disruption
- Limited visibility into asset health
After:
- Predictive maintenance at scale
- Planned interventions instead of failures
- Improved passenger experience
- Data-driven facilities operations
Why This Worked
- Focused on critical assets firstHighest operational impact systems instrumented early
- Used real failure dataModels trained on actual maintenance history, not assumptions
- Integrated into existing workflowsNo new system friction for maintenance teams
- Balanced prediction with actionabilityAlerts tied directly to work order creation
- Phased deployment strategyEnabled adoption without operational disruption
The Key Insight
Most maintenance costs are not from repairs. They are from:
- unplanned downtime
- emergency response
- operational disruption
Predictive maintenance shifts the equation:
failure becomes scheduled workcosts become predictableoperations become stable
Final Outcome
Aerolia transitioned from:
- reactive → predictive
- disruptive → controlled operations
- high-cost emergency fixes → planned interventions
Result: A facilities operation that is more reliable, more efficient, and significantly less expensive to run.
Optimising Large-Scale Infrastructure?
Intagleo Systems builds predictive maintenance platforms for complex environments, from airports to hospitals and smart cities.