IIoT-Based Machine Behavior Passport for Low-Cost Adaptive Condition Monitoring and Maintenance Recommendation

Industrial machines can be very important within manufacturing industries. However, unexpected failures can cause delays and increase maintenance costs. Traditional maintenance strategies can involve static threshold values which can cause issues with detecting early signs of abnormality as every machine will have differing characteristics. This paper describes an IIoT based Machine Behavior Passport which can be used to monitor the behaviour of machines in an adaptive and low cost manner. An ESP32 controller was used along with sensors to monitor the operational status of machines. During the learning phase, the system will record a machine's normal behaviour to create a specific baseline for that machine. The sensors can then be used to compare this baseline with real time data to detect any abnormality. Multiple sensors will be used to calculate a degradation index and produce a Machine Health Score within the range 0-100. The system also aims to detect probable fault conditions and recommend how urgent a maintenance is. An OLED screen was used to display local monitoring data along with a buzzer and LEDs for alerts. The system can also use an MQTT based cloud connection to give remote monitoring capabilities. A Micro SD card can also be used to store data in case of any temporary Wi-Fi issues. An affordable solution to monitor machine conditions and predict maintenance requirements can be provided.