As manufacturers begin to see multiple changes across the manufacturing industry, such as the increasing talent shortage, the increased reliance on technology will become apparent — from automation to replace employee labor to the ability to identify and manage exceptions throughout a manufacturing process. Manufacturers will be adjusting long-standing methodologies to incorporate and compensate for these new influences with the use of technology, including Manufacturing Visualization or Enterprise Manufacturing Intelligence (EMI) software. EMI and comparable visualization methodologies will still play a vital role in providing real time information of operations in factories around the globe to business executives, process owners and plant management. What is changing is how these tools are used in combination with other tools that manage key performance indicators (KPI) and support continuous improvement efforts throughout the manufacturing process.
One key change in the manufacturing process will come in the form of visual analytics. Defined as information visualization that focuses on analytical reasoning, the concept of visual analytics is becoming a hot topic in the manufacturing industry. With the ability to process large amounts of Big Data, analyzing the process of an outcome will become more streamlined and will change the current visualization methodologies of manufacturing.
From the executive’s point of view
First, let’s examine an executive’s EMI view as a basic example of visual analytics. A manufacturing executive will review a set of visualized KPI’s and slice those KPI’s based on plant locations, stockkeeping units (SKU) being produced or other relevant factors. Executives will then turn to Big Data visual analytics if a large segment of historical data is needed, or if the company is extremely large and the current data being analyzed is large enough to warrant the application of Big Data. If the executive finds an anomaly, he or she will obtain information to understand three issues:
- What action is being taken to resolve the current issue?
- Why did this happen?
- What is being undertaken to ensure this does not occur in the future?
The first piece of information that focuses on KPI visualization is simple enough to obtain, but the second and third issues require connecting with multiple associates, typically through a manual process. The process then allows for the executive to make a correct decision in a potentially complex situation, depending upon the results. However, the approach to decision making needs to evolve with the process for analyzing data to correctly manage the outcomes that are discovered.
The need for visualization
The current generic approach to tackling visual analytics, as outlined above, follows a standard pattern: data is transformed by specific models, the output of the models are visualized and the knowledge about specific outcomes is obtained to make decisions. While this is a standard approach, it will require modification as the urgency to make decisions increases, which may bring about a new model for the usage of visual analytics. This possible new model targets historical data analysis so that users can understand historical business events and performance outcomes. Most importantly, this new model will incorporate when exceptions are detected, what actions are taken, or whether the process that was utilized to resolve the exception and outcome are acceptable or not. The reason behind this is the rise of continuous streaming analytics that identifies exceptions and can make decisions faster than a person could ever dream of, which allows for actions to be taken more quickly and for more value to be captured by the correct decision.
It may sound very counterintuitive but visualization for a manufacturer will be used to support decisions on how exceptions will be managed in the future. This is why incorporating and visualizing the process steps and the outcome of the followed process is crucial to the long-term success of a manufacturer. Soon enough, the information about how a manufacturer got to a specific outcome will become just as important as the outcome itself. However, using visual analytics goes one step deeper.
Visual analytics enables manufacturers to quickly identify best practices and drive future automated decisions throughout the entire enterprise by combining the outcomes and the processes used to obtain those outcomes. This will serve manufacturing in the future well as the current talent gap experienced by the industry today continues to become more severe in the future.
The future of manufacturing will see visual analytics being combined with continuous streaming analytics to support analysis of the manufacturer’s decision rather than simply the analysis directed towards finding an exception. The use of visual analytics in the future will require a change in the manufacturer’s current thought process on how to leverage visual analytics.
Sean Riley is Director of Industry Solutions – Manufacturing & Supply Chain at Software AG.