Start Small and Grow Your IIoT Predictive-Maintenance Solution

Manufacturers know that unscheduled downtime due to tool failure is costly. Industrial Internet of Things (IIoT) solutions can help predict the maintenance needs of factory tools before they fail, providing tool owners and managers with options that best fit their operations. But how do you know if the investment in an emerging technology like IIoT will result in measurable savings?

At Intel, we face the same challenges. We have used IIoT in our factories for years and believe that understanding the return on investment (ROI) for any solution before we commit to scale is critical to a successful deployment. When we set out to gain visibility into the health of our tool fan filter units (FFUs), we chose a single sensor analytic metric to get us started because it was small enough in scope and large enough in impact to demonstrate ROI. You can read more about our FFU IIoT solution in the white paper: Developing a Scalable Predictive-Maintenance Architecture.

Predictive Maintenance with IoT Gateways and Sensors

Failing tools and components can lead to manufacturing excursions, unscheduled downtime, higher maintenance costs, and longer cycle times. And many existing tools and components, such as pumps, are not equipped with sensors, making the process of detecting tool health a manual-intensive process.

At Intel, we recognized that the vibration of our FFUs could be a valuable indicator of their health, but we needed a simple solution that could measure vibration patterns with the potential to scale other components and tools in the future. Our solution was to install vibration sensors on our FFUs and collected data through an IoT Gateway. Using data analytics, we were able to determine baseline measures for healthy FFUs. With these baselines, we were then able to identify when FFUs that fell outside of the baselines were likely to fail.

With our FFU IIoT solution, we realized the following benefits:

  • We increased our FFU uptime by over 97 percent, improving our factory floor operations.
  • Detecting inconsistencies in FFU vibrations has decreased excursion.
  • Technicians and tool owners can more quickly identify changes in FFU health.
  • We reduced our unscheduled downtime by 300 percent over our previous manual inspection process by ordering parts ahead of time.

IIoT provides a view into tool health that can help predict maintenance needs. With Intel® Internet of Things (Intel® IoT) technology, manufacturers can collect and compare data from tools across the entire factory. Knowing when a tool is unhealthy can save manufacturers money, increase factory throughput, and improve product quality. Predicting tool failure cost-effectively is achievable, but any IIoT solution should begin with measurable success criteria. By understanding the value of what you want to collect and starting small with a single, definable process, an IIoT solution can demonstrate ROI and lay the foundation for measuring the health of other tools and components throughout the factory.

Read the IT@Intel White Paper Developing a Scalable Predictive-Maintenance Architecture.

Published on Categories Internet of Things, ManufacturingTags , , , ,
Robert Colby

About Robert Colby

Rob is a Principal Engineer in IT Infrastructure responsible for Network Manageability and IoT Infrastructure. Rob joined Intel in 1999 in the "Intel Online Services" group, moved to factory security architecture in 2003, and in 2010 Rob shifted to focus on Location/Proximity infrastructure where his deep experience with wireless sensor technologies has enabled IT to standardize on wireless telemetry and location sensor infrastructure. Recently, Rob has been driving IoT infrastructure standardization with the IoT TWG, defining how IoTG and IoT market products land within Intel to solve our business problems. Rob has been an Intel Patent committee member for five years and holds ten US patents himself. Rob lives in California with his wife and four children, and in his spare time he enjoys working on his HO scale model train.