IoT Based Predictive Maintenance Management of Medical Equipment.

Area of Science:

• Biomedical Engineering
• Healthcare Technology Management
• Predictive Analytics

Background:

• Healthcare industry growth driven by technological advancements.
• Increasing need for optimal medical equipment maintenance strategies.
• Limitations of traditional corrective maintenance in ensuring equipment performance and cost-efficiency.

Purpose of the Study:

• To present a Predictive Maintenance (PdM) approach for critical medical equipment failure diagnosis.
• To integrate physics of failure understanding, IoT data collection, and machine learning for health status prediction.
• To conduct an economic analysis validating the feasibility and efficiency of transforming maintenance strategies.

Main Methods:

• Utilized Internet of Things (IoT) for real-time parameter collection via wireless accelerometers.
• Employed machine learning, specifically Support Vector Machine (SVM), for fault detection using vibration signals.
• Conducted a case study on a Vitros-Immunoassay analyzer in a UAE hospital, focusing on metering arm belt slippage failure mode.

Main Results:

• The PdM approach successfully predicted equipment failures, specifically metering arm belt slippage.
• Achieved significant diagnostic and repair cost savings, up to 25%.
• Demonstrated a one-year investment payback period for the implemented PdM system.

Conclusions:

• The proposed PdM approach effectively enhances medical equipment maintenance by predicting failures.
• The integration of IoT and machine learning offers a scalable and economically viable solution for hospitals.
• This strategy supports improved patient care through reliable equipment performance and reduced operational costs.