Today's machine control: intelligent and capable of learning

Details: Released: 08.04.2020; Last updated: 09.04.2023; Hits: 6091

With Artificial intelligence (AI) and advanced analytics (AA) can Machine control today process data, learn from it and make autonomous decisions. This increases the availability of machines and systems. These become more efficient, more reliable and more productive overall. How to embed these intelligent technologies in the steering can deliver a new paradigm of operation, knows Nils Knepper from Mitsubishi Electric.

It’s not that long ago that technologies like PID control, model-based control, field-oriented control and Fuzzy logic were just hypothetical. They are now so deeply embedded in the control architecture that we don't even think about them anymore.

Graphic HMI operator terminals for industrial automation

At this rate, in a few years, we'll be using artificial intelligence (AI) and advanced analytics (AA) take for granted in the machine control. You will be a driving force for increased machine availability. For example, this will be even more effective Predictive Maintenance (predictive maintenance) than we already have today.

Machine control with AI and AA competence

Using AI and AA technologies, we can at Mitsubishi Electric Big data analytics by recording and analyzing machine conditions in real time. They monitor the current machine status, anticipate malfunctions and give immediate recommendations for action. Machine operators and maintenance staff can either react, or the machine controller itself initiates appropriate measures.

How to make a protective mask in your own production

One step further, AI technology spans the whole company. If there were, for example, a bottleneck in the supply chain, the machine would react itself and slow down production itself until the spare parts arrived. This can prevent the entire production line from stopping.

In the future, artificial intelligence will optimize productivity autonomous decisions to meet. Today, a machine works within defined performance limits and takes into account, for example, different loads, safety areas or speeds. AI, however, uses deeper learning algorithms within the machine control. With them, machines can reach their limits today and go beyond them. Productivity can be increased without compromising reliability and security.

Maize species and edge computing are already present

AI technologies are already leading to operational improvements in individual machine processes. One example comes from Mitsubishi Electric: The automation company has developed a diagnostic technology based on its AI technology "Maisart" (Mitsubishi Electric's AI creates the state-of-the-ART in technology).

Smart maintenance: what influence does Industry 4.0 have?

This was implemented in products like the Edge Computing Solution "Melipc" that uses machine learning to analyze collected data. From this, it generates a model of the operating states of the machine. This model detects anomalies in machine operation in real time and notifies maintenance personnel of impending problems at an early stage.

Even the intelligent one predictive maintenancefunction "Smartplus ”in the“ Melfa ”robot is an example of the use of artificial intelligence. The smartplus analyzes primary drive components based on the existing operating conditions and warns early of failure or wear. In this way, maintenance can be planned efficiently. But the technology offers even more: During the design phase of the applications, the lifespan of the robot can be predicted and the annual maintenance effort can be estimated using simulation. This allows developers to modify the operation of the robot to extend its lifespan.

PLC control: Platform for networking in automation

These two examples already lead to a significant improvement in machine availability and lower maintenance costs. However, these initial uses of AA and AI only indicate their potential.