Predictive maintenance (PdM) is a growing opportunity in facilities management, using real-time data and analytics to anticipate equipment issues before they result in failures. While still emerging in FM, small-scale pilot projects are showing promise, especially for organisations willing to experiment with IoT sensors and basic predictive tools.

This practical guide provides a checklist for implementing predictive maintenance, highlighting realistic starting points and actionable steps that align with current industry capabilities.

1. What is Predictive Maintenance in FM?

Predictive maintenance leverages data from sensors and historical trends to predict when equipment needs attention. Unlike traditional reactive maintenance or fixed schedules, PdM ensures servicing is done only when necessary, preventing downtime while optimising costs.

Realistic Benefits at This Stage:

1. Improved Equipment Monitoring: Basic IoT sensors provide continuous updates on equipment health.
2. Reduction in Reactive Repairs: Early detection of wear and tear minimises unexpected breakdowns.
3. Pilot-Scale ROI: Organisations can see measurable savings in small test areas before scaling up.

2. Common Challenges in Predictive Maintenance

Barriers to Early Adoption:

1. Cost Sensitivity: Initial investments in sensors and software may deter FM teams.
2. Lack of Technical Expertise: Many FM teams lack in-house AI or data analysis capabilities.
3. Data Limitations: Historical maintenance data might be incomplete or poorly formatted.

3. Step-by-Step Predictive Maintenance Checklist

Step 1: Choose a Test Area for Your Pilot

• Objective: Start small with a manageable subset of assets.
• Actions:
  1. Identify high-value or frequently failing equipment, such as HVAC units or elevators.
  2. Select a single building or floor where equipment issues have caused disruptions in the past.
  3. Define clear goals, such as reducing downtime by 10% within six months.

Realistic Example:

A medium-sized office decided to test PdM on two air conditioning units that frequently failed during summer. The goal was to detect anomalies early and prevent mid-season breakdowns.

Step 2: Install Basic IoT Sensors

• Objective: Collect real-time data from selected equipment.
• Actions:
  1. Partner with an affordable IoT vendor to install vibration or temperature sensors.
  2. Ensure sensors are easy to install and integrate with existing systems like BMS or CAFM.
  3. Test sensor accuracy and reliability for 2–4 weeks before full deployment.

Realistic Example:

A shopping centre added low-cost temperature sensors to its HVAC system to monitor cooling performance during peak hours. The data helped staff identify cooling inefficiencies before tenant complaints arose.

Step 3: Gather and Analyse Data

• Objective: Build a foundation for identifying patterns and making predictions.
• Actions:
  1. Use simple tools or spreadsheets to organise sensor data alongside maintenance logs.
  2. Clean the data by removing duplicates and standardising units (e.g., degrees Celsius for temperature).
  3. Look for basic trends, such as rising temperatures or increased vibration over time.

Realistic Example:

An NHS trust tracked weekly temperature fluctuations in its boiler system. Maintenance teams noticed a gradual increase in temperatures that signalled the need for servicing.

Step 4: Use Pre-Built Predictive Tools

• Objective: Take advantage of existing software to identify potential issues.
• Actions:
  1. Invest in an entry-level AI-powered tool compatible with your sensors.
  2. Use dashboards to visualise data trends and flag anomalies.
  3. Validate the tool’s accuracy by comparing predictions with real-world outcomes during a 3-month trial.

Realistic Example:

A local council used a simple, subscription-based platform that visualised sensor data from its water pumps. Alerts for unusual vibrations helped reduce emergency repairs by 15%.

Step 5: Automate Alerts and Responses

• Objective: Enable immediate action when anomalies are detected.
• Actions:
  1. Configure alerts for key thresholds, such as temperature exceeding safe levels.
  2. Use email or mobile notifications to inform relevant team members.
  3. Integrate alerts with CAFM platforms to automatically create work orders.

Realistic Example:

A retail store chain received automated alerts when refrigeration unit temperatures exceeded acceptable levels, enabling quick adjustments to prevent food spoilage.

Step 6: Measure Success and Refine

• Objective: Evaluate pilot results and improve processes.
• Actions:
  1. Track key metrics like downtime, maintenance costs, and repair frequency before and after PdM.
  2. Gather feedback from maintenance teams to understand operational challenges.
  3. Adjust sensor thresholds and software settings based on findings.

Realistic Example:

A leisure centre saw a 20% reduction in maintenance costs after refining vibration thresholds for its pool pumps during the pilot phase.

4. Practical KPIs for Predictive Maintenance

• Downtime Reduction: Track the percentage decrease in equipment downtime during the pilot.
• Emergency Repairs: Compare the number of reactive repairs before and after implementation.
• Maintenance Cost Savings: Calculate the reduction in labour and material costs.
• Prediction Accuracy: Measure how often alerts correctly identified issues.
• Pilot ROI: Estimate savings against the initial investment in sensors and software.

5. Addressing Common Predictive Maintenance Barriers

Barrier 1: Budget Constraints

• Solution: Start with affordable sensors and scale up based on pilot results.
• Example: Use a £500 sensor kit for basic temperature monitoring rather than investing in advanced, high-cost systems.

Barrier 2: Data Quality Issues

• Solution: Focus on simple, clean datasets during the pilot phase.
• Example: Use only the most relevant maintenance logs for the assets being monitored.

Barrier 3: Lack of Expertise

• Solution: Partner with vendors who provide training and support for sensor installation and software use.
• Example: A small museum worked with an IoT provider to set up and interpret vibration sensor data for its HVAC units.

6. Scaling Predictive Maintenance After a Successful Pilot

Once a pilot demonstrates measurable benefits, consider expanding the programme to other assets and locations.

Steps for Scaling:

1. Expand Asset Coverage: Add lower-priority equipment like lighting systems or auxiliary generators.
2. Integrate Additional Data Sources: Incorporate occupancy patterns or weather data to enhance predictions.
3. Standardise Organisation-Wide Processes: Train all teams to follow consistent PdM workflows.

Key Takeaways

1. Start Small and Realistic: Focus on affordable pilots for high-priority equipment.
2. Use Simple Tools: Basic IoT sensors and dashboards can yield valuable insights.
3. Focus on Measurable Outcomes: Track downtime, costs, and prediction accuracy to build confidence in PdM.
4. Refine and Scale Gradually: Use pilot results to justify broader adoption across the organisation.

Next Steps with Baachu Rain

At Baachu Rain, we specialise in helping FM teams implement practical predictive maintenance strategies. Whether you’re starting with a small pilot or scaling organisation-wide, we provide the tools and support you need.

• Explore Our Data Integration Guide
• Learn More About Compliance Monitoring Tools

📧 Contact Us: hello@baachu.com

📞 Call Us: +44 203 574 8855

Let’s help you build a predictive maintenance programme that works for your facilities and budget.