Already in use: edge computing, AI and blockchain

An essential component of factory automation is the use of artificial intelligence. Today, a wide variety of AI applications are already part of everyday industrial life: speech and gesture recognition in human-machine interfaces, industrial image processing, the detection of environments with cameras, laser beams or X-rays in autonomous robots, all the way to predictive maintenance.
 

Machine Vision with AI

The use of artificial intelligence offers precisely the flexibility and performance that highly automated manufacturing needs. Take machine vision, for example: conventional, rule-based machine vision only recognises clearly quantifiable, unambiguous and very specific attributes (e.g. different shapes or a horizontal scratch on a surface). AI-based machine vision, on the other hand, can also deliver results for non-quantifiable attributes, recognise defects even under varying environmental conditions (lighting, backgrounds) and deal flexibly with product variations and defect shapes (e.g. dents, colours).
 

More intuitive operation with AI

AI is also changing the way machines and systems are operated. Natural Language Processing (NLP) is increasingly complementing conventional human-machine interfaces. With the help of AI algorithms, NLP processes the spoken word of the machine operator or robot programmer into machine text. Gesture control, in which camera systems and sensors recognise the operator's movements and translate them into machine instructions, is also becoming increasingly common in factory automation. The advantage of these AI-supported human-machine interfaces is that machines and systems can be operated more intuitively. In addition, classic input technologies such as keyboards are not always problem-free in manufacturing - for example, when maximum hygiene is required, the operator must wear work gloves or simply must have his hands free to carry out his work.
 

Real-time processing thanks to edge computing

Thanks to special, high-performance chips, AI applications are no longer only possible in the data centre, but also directly at the location where the data is generated. This edge computing enables real-time processing and thus the reaction times that the fast processes in factory automation require. For example, sensors for reading barcodes can recognise and decode codes on the object faster and more reliably thanks to integrated AI processors and thus trigger actions in the conveyor lines in good time.
 

From factory automation to business automation

Increasingly powerful automation not only has an impact on the product and its production, but also increasingly on business models. This is because comprehensive real-time data is available, which offers suppliers and customers basic parameters for a wide range of services. This begins with the monetisation of predictive maintenance services, for example, and extends to completely new sales concepts such as pay-per-use, where the operator pays for systems and machines according to the time used or the number of units produced.

In combination, AI and blockchain open up many new possibilities in the context of factory automation (Source: LeewayHertz)

If you think about automation beyond manufacturing, a technology previously known mainly from the financial sector comes into focus: the blockchain. Manufacturing companies exchange information about raw materials, production or logistics with various stakeholders. For each exchange, a financial, contractual and trust-based relationship between two parties is a prerequisite at all times. With blockchain technology, trust can be ensured within the supply chain or value chain and thus securely interact with each other without the control of a third party. In the context of factory automation, for example, storage costs and throughput times can be reduced, overcapacities can be distributed, and conditions or prices can be automated.