IoT Sensors for Predictive Maintenance: Implementation Guide

Predictive maintenance promises to catch equipment failures before they happen. IoT sensors make that promise achievable—but implementation matters more than technology selection.

This guide walks through practical IoT sensor deployment for facility maintenance teams, from sensor selection to CMMS integration.

The Business Case for IoT Sensors

The math is straightforward:

Reactive maintenance cost: Emergency repair + downtime + expedited parts + overtime labor

Preventive maintenance cost: Scheduled service at fixed intervals (often too early or too late)

Predictive maintenance cost: Service only when sensor data indicates degradation

Studies show predictive maintenance reduces:

• Unplanned downtime by 30-50%
• Maintenance costs by 10-40%
• Equipment failures by 70-75%

The catch? You need sensors, connectivity, and integration to make it work.

Sensor Types for Facility Maintenance

Vibration Sensors

Best for: Rotating equipment—motors, pumps, fans, compressors, generators

What they detect:

• Bearing wear (increasing vibration amplitude)
• Imbalance (specific frequency patterns)
• Misalignment (directional vibration)
• Looseness (irregular patterns)

Placement: Mount on bearing housings or motor frames. Ensure solid contact—loose mounting gives false readings.

Alert thresholds: Establish baseline during normal operation. Alert when readings exceed baseline by 50-100%.

Cost range: $100-400 per sensor

Temperature Sensors

Best for: Electrical systems, bearings, HVAC equipment, motors

What they detect:

• Overloaded circuits (rising temperature)
• Bearing failure (friction heat)
• Insulation breakdown (hot spots)
• Refrigerant issues (abnormal temps)

Placement: On bearing surfaces, near electrical connections, on motor housings, in electrical panels.

Alert thresholds: Set based on equipment ratings. Typical alerts at 80% of maximum rated temperature.

Cost range: $50-200 per sensor

Current/Power Sensors

Best for: Electric motors, pumps, HVAC equipment

What they detect:

• Motor degradation (increasing current draw)
• Mechanical load changes (power anomalies)
• Phase imbalance (electrical issues)
• Efficiency loss (power factor changes)

Placement: On power feeds to monitored equipment. May require electrician for installation.

Alert thresholds: Alert when current exceeds nameplate rating or deviates from baseline by 15-20%.

Cost range: $100-300 per sensor

Environmental Sensors

Best for: HVAC monitoring, indoor air quality, critical spaces

What they measure:

• Temperature and humidity
• CO2 levels
• Differential pressure
• Air quality (particulates)

Placement: At return air ducts (system performance), in occupied spaces (comfort), in critical areas (server rooms, labs).

Alert thresholds: Based on ASHRAE standards and space requirements.

Cost range: $50-200 per sensor

Leak Detection Sensors

Best for: Mechanical rooms, water heaters, under sinks, data centers

What they detect:

• Water presence on floors
• Condensate overflow
• Pipe leaks
• Flooding conditions

Placement: Low points where water would collect, under equipment prone to leaks.

Alert thresholds: Any water detection triggers immediate alert.

Cost range: $30-100 per sensor

Connectivity Options

LoRaWAN

Best for: Large facilities, multi-building campuses, rural locations

Advantages:

• Long range (2-10 km)
• Low power (battery life 5-10 years)
• Penetrates walls and floors well
• Low cost per sensor

Requirements: LoRaWAN gateway on-site, cloud or on-premise network server

Cost: Gateway $200-1,000, sensor modules $30-100

WiFi

Best for: Office environments with existing robust WiFi

Advantages:

• Uses existing infrastructure
• High bandwidth for data-rich sensors
• Familiar technology

Disadvantages:

• Limited range
• Poor penetration in industrial environments
• Higher power consumption
• Security considerations

Cost: WiFi sensors $50-300

Cellular (LTE-M/NB-IoT)

Best for: Remote locations, facilities without network infrastructure

Advantages:

• No local infrastructure required
• Works anywhere with cellular coverage
• Good for isolated assets

Disadvantages:

• Ongoing cellular data costs
• May have coverage gaps indoors

Cost: Cellular sensors $100-400, plus monthly data fees

Recommendation

For most facility deployments, LoRaWAN offers the best combination of range, battery life, and cost. A single gateway can cover an entire building or campus, and sensors last years without battery replacement.

Implementation Roadmap

Phase 1: Identify Critical Assets (Week 1-2)

Not every piece of equipment needs sensors. Focus on:

High criticality:

• Safety systems (fire pumps, emergency generators)
• Production-critical equipment
• Equipment with high failure consequences

High failure cost:

• Expensive to repair
• Long lead time for parts
• Causes significant downtime

Predictable failure modes:

• Rotating equipment (vibration patterns)
• Equipment with known wear items
• Systems with temperature-related failures

Create a prioritized list of 10-20 assets for initial deployment.

Phase 2: Select Sensors and Connectivity (Week 2-3)

For each asset, determine:

• What failure modes to monitor
• Which sensor types address those modes
• Optimal sensor placement
• Connectivity requirements

Map out gateway placement if using LoRaWAN—typically one per building or floor depending on construction.

Phase 3: Baseline Data Collection (Week 3-6)

Install sensors and collect data for 2-4 weeks before setting alerts.

Why baseline matters:

• Equipment has normal variation
• External factors affect readings (weather, load)
• Alert thresholds need to be calibrated to your specific assets

Run equipment through normal operating conditions during this period.

Phase 4: Configure Alerts and Integration (Week 6-8)

Based on baseline data, set:

Warning thresholds: Elevated readings that warrant monitoring (email notification)

Critical thresholds: Readings indicating imminent failure (immediate alert + auto work order)

Integration points:

• CMMS API for work order creation
• Dashboard for real-time monitoring
• Notification channels (SMS, email, app)

Phase 5: Continuous Improvement (Ongoing)

Weekly: Review alert accuracy—too many false positives? Adjust thresholds.

Monthly: Correlate sensor data with actual failures—did sensors predict them?

Quarterly: Evaluate expansion to additional assets based on ROI.

CMMS Integration

The real value of IoT sensors comes from integration with your maintenance workflow.

Automatic Work Order Creation

Configure rules:

When: Vibration sensor on AHU-01 exceeds 8mm/s Then: Create high-priority work order: “AHU-01 vibration anomaly detected—inspect bearings” Assign: HVAC team Include: Link to sensor dashboard, historical readings, asset maintenance history

Contextual Information

Work orders generated from sensor alerts should include:

• Current sensor readings
• Historical trend showing degradation
• Similar past issues and resolutions
• Relevant documentation

Feedback Loop

When technicians resolve sensor-generated work orders:

• Was the alert accurate?
• What was actually wrong?
• Should thresholds be adjusted?

This data improves prediction accuracy over time.

Cost-Benefit Analysis

Sample Deployment: 50 Sensors

Initial Costs:

Ongoing Costs:

Expected Benefits

Avoided emergency repairs: If sensors prevent 3 emergency HVAC callouts per year at $2,000 each = $6,000

Reduced downtime: If sensors prevent 2 major failures causing 8 hours downtime each at $500/hour = $8,000

Extended equipment life: If predictive maintenance extends asset life by 2 years on a $50,000 asset = $5,000/year equivalent

Total annual benefit: $19,000+

ROI: First-year payback with ongoing positive returns

Common Implementation Mistakes

1. Starting Too Big

Deploying 500 sensors across an entire portfolio before proving value on 20 assets.

Better approach: Start with 10-20 critical assets. Prove ROI. Expand based on results.

2. Ignoring Baseline Period

Setting arbitrary thresholds without understanding normal equipment behavior.

Better approach: Collect 2-4 weeks of baseline data before configuring alerts.

3. Alert Fatigue

Setting thresholds too sensitive, generating dozens of daily alerts that get ignored.

Better approach: Tune thresholds to minimize false positives. Every alert should warrant action.

4. No Integration

Sensors feed a dashboard that nobody watches. No connection to maintenance workflow.

Better approach: Integrate sensors with CMMS for automatic work order creation.

5. Missing the Human Element

Relying entirely on sensors without technician input on equipment condition.

Better approach: Sensors augment—not replace—technician observations and expertise.

The Bottom Line

IoT sensors for predictive maintenance aren’t science fiction—they’re proven technology that most facility teams can implement with modest investment.

Start with critical assets where failure costs are high. Choose appropriate sensors for the failure modes you’re monitoring. Integrate with your CMMS so insights drive action.

The facilities that master predictive maintenance will operate with less downtime, lower costs, and fewer emergencies than those stuck in reactive mode.

Ready to implement IoT-based predictive maintenance? See how Infodeck’s IoT platform connects sensors to your CMMS with native LoRaWAN support, automatic work order creation, and real-time monitoring dashboards.