In the past, overall equipment effectiveness (OEE) programs could jolt operators when management wanted to focus on how to improve plant utilization rates by examining machine performance. A lack of standardization on OEE metrics within plants or an enterprise is well documented, but as digitalization and remote monitoring become foundational, measurement processes are improving.
One company adopting advanced analytics and standardized processes to measure OEE across its plants isIllovo Sugar, based in Mount Edgecomb, South Africa. The company recently implemented an OEE program to measure the performance of vertical form/fill/seal (vf/f/s) machines across five sugar plants.
“One of our business strategies was to shift from bulk product to smaller stock keeping units (SKU),” says Lloyd Melrose, group C&I engineer at Illovo Sugar, during the 2018 PI conference, sponsored byOSIsoftTechnology. “We wanted to increase our customer base and needed to improve the performance of our form/fill/seal machines. So, we needed to measure, analyze, improve and control [the machines].”
Illovo has 49 vf/f/s machines across its plants and wanted to aggregate the data with OSIsoft’s suite of technology solutions, including Asset Framework (AF) servers, PI Vision, PI Historian, PI Manual Logger and PI Datalink Visualization tools. PI Asset Framework is a single repository for asset-centric models, hierarchies, objects and equipment. This platform integrates, contextualizes and analyzes data from multiple sources, including one or more PI Data Archives or external relational databases.
The data collection at Illovo starts Siemens S7-1200 programmable logic controllers at the machine level, sending status, bags per hour (bph), bags per minute (bpm), bag type and current shift status to an OPC server/client and then on to an Asset Framework Server for OEE calculations.
These machine calculations include hourly availability, hourly performance, hourly OEE, hourly quality, bpm, current bag type and current shift.
”使用多个报告费尔ters enables better contextualization of data and allows a comparison of machine performance between shifts,” says Melrose. “For example, we can see the performance per shift and can mark human (operator) deficiencies for training purposes and patterns.”
Reporting templates
Illovo configured the reporting for each packaging station using element templates within the AF hierarchy. A key element for reporting was the reason codes for the downtime. To apply reason codes, Illovo used the AF server and event frames to capture downtime events and reason codes from a predefined reason tree. The PI platform has a feature called event frames that captures critical event contexts. The event frames permit customers to obtain a name, start time, end time and a series of related information (event attributes) that are useful for analysis within the PI System.
In the early part of this initiative, Illovo focused on total downtime duration per reason code as well as the total counts or downtime events per specific reason code, according to the company. The human machine interface (HMI) reason code capture was done through a local HMI to each machine.
Illovo did some preliminary calculations locally and displayed the downtime information in real time. “This became very important for getting competition between shifts and finding improvement by providing real-time information to those operators,” says Melrose.
The project team comprised a continuous improvement manager, Melrose and four other members that cover different areas of support within the enterprise. “We managed to implement it within four months, and we saw benefits in about three months,” states Melrose.
