OEE's dark little secret

Figure 1: Normally a substantial amount of additional data is required to calculate OEE using the traditional calculation methodologies, resulting in costly integration efforts with a factory’s ERP and Planning Tool to automate. Above, we can see an example of how state-of-the-art tools can intelligently infer key parameters from existing machine data, without the need to integrate multiple systems. In the top diagram, we can see how the system automatically identified planed productive time (PT) and non-productive time (NPT). In the bottom diagram, we can see where AI has automatically recognised change overs on the line, with each product illustrated in a different colour. This also has the added benefit that it eliminates a lot of human intervention and unnecessary manual data entries, reducing the possibility of human error.

Figure 2: The iTAC.Asset.Analyzer’s system automatically detects anomalous system behaviour by monitoring measurement data and detecting outliers in the data set in real time. These can be used to set alarms or trigger workflows if certain thresholds are reached.

Figure 3: An example of an automated message sent by the iTAC.Asset.Analyzer to the process engineering team on shift duty to investigate a problem at the Conformal Coating Inspection (CCI) machine after it identified an unusually high cluster of quality problems. Here an external workflow is triggered, whereby the team is informed of anomalous behaviour and impeding situation via MS Teams.

Figure 4: Modern tools understand not only the dynamic of each machine, but also how the behaviour of all machines on the line are affected by each other in certain situations like breakdowns. Above we see how the iTAC.Asset.Analyzer uses digital twin models to augment the real-time data stream and provide very accurate state representatives of each machine and the overall status of the line. These give production staff a deeper understanding of the root causes behind production outages and reduced output. 

Figure 5: The iTAC.Asset.Analyzer’s real-time tracking uses an efficiency matrix to show system users how efficiently individual resources in a production line are utilized, and how much they contribute to total flow efficiency. The Perfect State is machines running both efficiently and lean, shown in the top left corner of the matrix. Islands of Efficiency, or machines running efficiently in isolation but not in relation to other machines in the line, are shown in the top right corner. Efficient Oceans, or machines running synchronously with each other but being underutilized, are in the bottom right corner. Finally, machines in the Wasteland category (bottom left corner) are not resource or flow efficient and make up the most undesirable scenario. In this example we see that the bottleneck machine (pnp_2) is running below maximum resource efficiency. This situation is not ideal, and management can immediately recognize that there are huge potential productivity gains to be made across the entire line if the utilization of this one machine can be improved. 

An innovative system like the iTAC.Asset.Analyzer’s AI Playbook captures human expertise, documents human actions to identify their root causes and even provides actionable recommendations. This helps teams learn from past challenges, develop proactive solutions, and continuously improve their performance. Fully integrated Connected Worker Solutions such as these are empowering next-generation AIs to deliver unprecedented efficiency gains.

Figure 6: An example menu from the iTAC.Asset.Analyzer that allows operators to add additional comments information during the problem resolution process, allowing their process knowledge and factory expertise to merge with real-time data streams.