According to the industry association Bitkom, two thirds of German companies see artificial intelligence as an important technology of the future. However, there is often still a lack of well thought-out and comprehensive implementation strategies. However, there are many reasons why it is important for companies to get to know innovative technologies and approaches, to determine their purpose and benefits for their own company and to question them: Which areas can be streamlined? How can AI and factory automation support employees in their work? When selecting new machines, for example, care should be taken to ensure that they have the necessary functions to make data available for AI purposes in the future.

The interaction between man and machine

The AI controller from Omron provides support directly at the process in order to use AI in machine-oriented real time.

© Omron

A vivid example of AI in an industrial environment is the 'Forpheus' table tennis robot from Omron. The main aim of this technology (which has already been presented at many industry meetings and trade fairs) is to demonstrate the harmonious interaction between man and machine. Forpheus can inspire smart industrial automation and production processes of the future, as the table tennis trainer impressively demonstrates how human capabilities can be maximized and supported with the help of intelligent technology. AI, robotics and sensor technology are used together to make people's work easier, predict situations and enable them to react to challenges at an early stage. This combination of different solutions can also be used in production to achieve 'harmonization'.

AI can make machines and people smarter: The experienced employee trains the machine and the machine trains the unskilled employee - as with the Forpheus robot. The table tennis game serves as an example for many work steps that are conceivable in production or packaging, warehousing or quality control. Omron, for example, is researching AI-controlled machines that prompt operators to assemble products and comprehensively document work steps in order to determine the best methods and work techniques. AI can also be used to check which actions an operator should perform on the machine in order to avoid errors. If the operator's hands move in the wrong direction, for example, an alarm is triggered and a recommendation is displayed.

Skills shortage requires new strategies

Why it is so essential for companies to consider new strategies and focus on innovative solutions is demonstrated by the shortage of skilled workers, which poses enormous challenges for companies in a wide range of industries. The German Chamber of Industry and Commerce (DIHK) estimates that there are around two million vacant jobs in Germany - and the trend is rising. According to the DIHK, the proportion of companies that expect their employees to take on additional work is highest in industry, at 66%. It is therefore important to find ways to support human skills through the use of AI and employee-oriented automation technologies. But what does this look like in practice?

Real system autonomy

One consideration to start with: controlling AI-based systems requires a lot of expertise, so they are currently mainly found in niche applications. Developers are therefore increasingly focusing on systems that can independently find out why they have stopped, for example, or why there is a problem. While sensors enable operators to find out that a cartoning machine is missing blanks, for example, they do not indicate whether the paper density is outside the tolerance range. In other words, sensors alone are not enough. In order to achieve true system autonomy, supporting artificial intelligence is required, as machines can perform sophisticated analyses with the help of intelligent algorithms. For industrial applications, it is advisable to start small and with a very specific problem. The relevant data must then be collected and stored synchronously. No information should be lost in the process. Benefits arise from the subsequent data analysis.

AI is not a panacea

The use of AI in the sealing process makes it possible to minimize faulty seals.

© Omron

AI-supported calculations can be carried out faster and more cheaply than before thanks to continuous improvements in mobile and memory technology and processing speed. However, AI is not a panacea and it does not make sense to rush to technical solutions: If, for example, a section of a conveyor belt is damaged and bent, this problem can be identified and solved mechanically. AI offers added value, particularly for less obvious and individual challenges. One example of this can be found in the food industry, where Omron is currently working with a customer to improve seal quality and seal integrity. The use of AI in the sealing process makes it possible to extend the shelf life of the product by several days and minimize faulty seals.

Seamless data integration

The "Forpheus" table tennis trainer impressively demonstrates how human skills can be maximized and supported with the help of intelligent technology.

© Omron

Another example comes from the automotive sector: Omron Automotive Electronics Italy (A.E.I.) - a company that produces over 30 million components a year for the global automotive market - uses data analysis for the intelligent real-time management of process-based errors. Initially, the production line for G8HN power relays was upgraded. The 'i-Belt Data Service' used is responsible for AI-related data acquisition and processing in the machine. This allows difficulties in production and administration to be linked with production data. The data is converted into useful information that can be used to implement improvements in practice and overcome production challenges. In order to use AI in machine-oriented real time, Omron's AI controller provides support directly at the process. This AI solution, which works 'at the edge', is based on the 'Sysmac NY5' IPC and the 'NX7' CPU. As the controller is seamlessly embedded in the 'Sysmac' production control platform, there are no interfaces or data breaks to overcome, which improves efficiency. The AI controller recognizes anomalies in the process very accurately and provides immediate feedback.

Robots learn optimal assembly

AI in production can also be easily integrated into assembly: On the one hand, robots and cobots learn from humans and, on the other, teach them to further develop their own skills - similar to the table tennis robot Forpheus. The basis for this is the 'Sensing & Control + Think' technology: the robot learns how to carry out assembly tasks by following hand movements and learning from mistakes made by the human operator. The knowledge gained in this way enables the machine to develop an optimal strategy for assembling a product. The next step is to create a digital twin of the robot to train other operators in a virtual environment. This teamwork between man and machine ensures continuous performance improvement. The CO2 production at production sites in different countries can also be tracked in real time via a live data space connection. In this way, the energy efficiency of different sites can be improved by comparing the CO2 emissions from the manufacture of similar products.

Cell Line Control Systems (CLCS) are another example of a technology that will optimize and accelerate production processes. They are equipped with a production control mechanism and an information platform to produce defect-free products as efficiently as possible in multi-product manufacturing with variable volumes. Such CLCSs also include interactive manufacturing systems in which a cobot is placed next to an operator to work synchronously with humans in a work area.

Gradually scaling optimization potential

The author: Dr. Klaus Kluger is Sales Director Central and Eastern Europe at Omron in Langenfeld.

© Omron

Skills shortages, sustainability efforts, supply chain bottlenecks and competitive pressure are just some of the hurdles that industrial companies of all sizes currently have to overcome. Decision-makers are therefore increasingly challenged to maximize human capabilities through user-centric automation technologies and AI. The development of a harmonized automation concept helps to solve this problem. Future-proof approaches combine automation technologies such as robotics, control, sensors and image processing with technologies such as AI, AR and 5G to develop solutions for sustainable manufacturing. To do this, companies must now analyze and question their processes: Where is there potential for optimization and where is automation paired with AI suitable for the individual problem?