Predictive Advantage

RISK-BASED MAINTENANCE METHODOLOGY

Our six-step BFM analysis framework created a comprehensive risk profile for every asset:

• Assessment of business needs and consequence of failure
  This stage required deep consultation with the customer to understand business priorities. For example, a pump removing water from a room carries a different criticality than an HVAC unit in a server room, even if both are technically important.
• Preparation of functional block diagrams and assessment of functional resilience
  Through functional block diagrams and visualisation, the team could identify dependencies and understand cascading failure risks. An asset that will normally score highly might receive a lower score if redundancy exists elsewhere in the system, allowing smooth degradation over time, rather than complete failure.
• Conducting condition assessments to analyse past failures and identify any recurring failures
  We assessed both the current condition and past asset failure data. Recurring failures prompted a deeper investigation into root causes and informed scoring decisions.
• Calculating the BFM risk score from the previous three steps
  We used a scoring table out of 100 to assess the likelihood of failure. These scores then determined the next step of either reducing operations and saving costs, reallocating actions, or enhancing approaches based on condition-based monitoring or predictive maintenance strategies available.
• Using the score to determine whether to reduce maintenance, continue with SFG20 optimal maintenance, or enhance with condition-based monitoring
  Assets scoring very low could afford reduced maintenance, indicating they had not failed, maintained good status, and were not problematic. Removing these maintenance activities from schedules generated cost savings. Assets in the optimal range could continue with SFG20 standard maintenance. Assets scoring in the optimal-plus range, those failing frequently or critical to operations, became candidates for enhanced maintenance approaches.
• Implementing these maintenance changes and periodically reviewing their effectiveness
  For the top-scoring assets, we considered enhanced monitoring through IoT sensors or additional BMS sensors to enable the shift towards condition-based and predictive maintenance. The goal was to identify the optimal intervention spot on the Potential to Functional Failure, the P-F, curve. The curve represents the period of graceful degradation during which maintenance intervention is most effective before catastrophic failure can occur. But this implementation is only as effective as the assets it is applied to. Some assets, like lightbulbs, provide no warning before failure, and no amount of sensor implementation would predict their failure. So, selecting which assets would benefit from enhanced monitoring and selecting specific strategies for each was key.