Vibration analytics-based system predicts IIoT faults

Article By : Harman

The Harman Quick Predict is an end-to-end warning solution for industrial Internet of Things based on vibration analytical algorithm.

Harman and Intel have announced an end-to-end industrial IoT solution that enables early detection of problems with rotating equipment in industrial settings—the Harman Quick Predict. Harman’s engineering services team developed Quick Predict based on a vibration analytical algorithm developed by Intel.

 

Figure 1: Harman Quick Predict is an Industrial IoT solution which has been developed based on Intel’s intellectual property.
 

Harman Quick Predict generates predictions based on real-time information instead of historical data collected over time. Using machine learning algorithms, the solution enables prediction of potential failure based on capture and analysis of abnormal vibration patterns at the gateway that are detected and flagged. The Harman Quick Predict collects the high-resolution vibration data needed to detect problems early and provides learning analytics that help map abnormal vibration and rotation speed patterns to associated failure mechanisms.

Rotating equipment maintenance is expensive and resource intensive. Even with full spare parts and machinery in place, a pump failure on a line can cause costly production delays leading to emergency work orders and hurried scheduling of maintenance crews. Spot manual vibration readings collected under preventative maintenance programs on a weekly or monthly basis by technicians simply do not provide the data needed to identify all problems early enough to allow for planned repair.

An IoT technology originally intended for Intel fabrication plants to improve maintenance and uptime, the Harman Quick Predict is an integrated solution that will increase equipment uptime, decrease spare parts costs and optimise the use of workforce by reducing the number of emergency repairs, according to the company.