iFactory PHM

Prognostics and Health Management (PHM) is a cutting-edge solution for intelligent factories. This advanced system excels in predictive maintenance, employing real-time data analysis to detect potential risks and predict equipment health. By anticipating potential machine issues, EquipPHM enables users to implement proactive maintenance measures before problems escalate. The result is swift identification of machine health status and timely alerts, effectively preventing unexpected downtimes. With a keen focus on predictive capabilities, EquipPHM optimizes labor and spare parts management, reducing repair times and boosting operational resilience.

Features

1. AI-powered health status forecasting

• Pre-training using vibration and temperature data.
• Retrain IoT device directly on the machine.
• Obtain inference results specific to the machine
• PHM AI model references machine health pattern as a golden sample for evaluating the machine's operating status.

2. Early failure risk detection for advanced machine handling

• Real-time identification of unexpected failure risks.
• Prediction of machine health scores for early-stage failure detection over 7 days.
• Visualization of analytical results through a dashboard.

3. Cloud and On-prime services available

• Light-weight AI inference task embedded in calculation server.
• Uses SaaS to eliminate environment maintenance efforts to ensure the solution stays up-to-dat

4. Apply AI results for value-added benefits

• Easily integrate AI results with existing services using open API.
• Extensive applications available on WISE-IoT

5. Adapted for different types of rotating machines

• Designed for motor-driven rotating machines which are ISO 10816 applicable
• No need for model customization.
• Wireless sensor monitoring for fan, pump, cooling tower, compressor, and other factory facilities.

6. Agile application

• Install the sensor and initiate inference tasks through a simple process.
• Collect only data from healthy machines, excluding any related to machine failures.
• Retrieve inference data after accumulating 200 data points.

System architecture