Not so long ago, technologies such as model-based control, PID control, field-oriented control and fuzzy logic were purely hypothetical. Today, they are so deeply embedded in control architectures that we no longer even think about them.

How will we think about advanced analytics (AA) and artificial intelligence (AI) in machine control systems in a few years' time? In case of doubt, quite naturally. Because they will be a driving force for increased machine availability, for example by enabling even more effective predictive maintenance than is already possible today.

Analyze machine statuses in real time

AA and AI technologies make it possible to record and analyze various machine states in real time using big data analyses. They monitor the current machine status, detect impending malfunctions and provide immediate recommendations for action. The machine operator or maintenance service can react, or the system will initiate corrective measures independently.

What happens when AI technology is rolled out across an entire company? A scenario for the supply chain would be exciting. There are bottlenecks in delivery. The machine reacts and slows down production on its own until the replacement components arrive. This can prevent the entire production line from coming to a standstill.

In the future, AI will make autonomous decisions to optimize productivity. Currently, a machine is designed to work within defined performance limits - perhaps to take into account different loads or speeds or safety areas. AI technology, however, uses deeper learning algorithms within the control system. These could make it possible to drive machines to their current limits and beyond. And thus significantly increase productivity without compromising reliability or safety.

AI improves processes

We are already seeing how the application of AI principles to individual machine processes can lead to operational improvements. For example, Mitsubishi Electric has developed a diagnostic technology based on its AI technology called Maisart (Mitsubishi Electric's AI creates the State-of-the-Art in technology). Embedded in products such as Mitsubishi Electric's Melpic edge computing solution, it uses machine learning to analyze collected data to generate a model of the machine's operating conditions. This model can detect anomalies in the machine's operation in real time, providing early warning of impending problems so that maintenance personnel can take immediate action.

Another example of the use of AI is the intelligent predictive maintenance function of the Melfa robots. The Melfa SmartPlus function can be applied to Mitsubishi Electric robots. It accurately analyzes primary drive components according to actual operating conditions and provides early warning of part failure or wear. This reduces downtime and allows an efficient maintenance schedule to be planned. In addition, the technology provides simulation capabilities during the design phase of applications to predict the robot's service life and estimate annual maintenance costs. This allows engineers to modify the operation of the robot to extend its service life.

These two examples are already leading to a significant improvement in machine availability and a reduction in maintenance costs. And that is just the beginning. Because they merely point to the potential of AA and AI.

Nils Knepper, Senior Product Manager Modular PLC Central Europe for Industrial Automation Systems, Mitsubishi Electric Europe B.V.