One focus of the Mejoin project, which is funded by the German Federal Ministry of Education and Research (BMBF), is the exchange of technology and market expertise with a focus on artificial intelligence (AI) and the Internet of Things (IoT). One of the main questions addressed by the Mejoin project is the extent to which the participating companies have integrated artificial intelligence into their business model and how they can achieve economic benefits by using new technologies in the future.

Support from robots

Some of the companies involved in Mejoin are active in robot-related areas. Robots are increasingly being used to optimize production and for particularly dangerous or physically demanding tasks. Through the use of AI, their area of application is expanding to include constantly changing components and situations in production. Various AI learning techniques such as simulation learning and deep grasping shorten the learning time of robots and improve their performance in an industrial environment. The use of cameras and image data processing using artificial intelligence can significantly improve the learning and performance time of assembly robots.

The trend towards collaborative robots (cobots) in particular enables the use of robots in SMEs. They offer a new platform for the application of machine learning processes in these companies. The use of machine vision technology is particularly suitable for this - after all, the precision of image recognition increased from around 72% in 2010 to over 96% in 2015. This means that in many areas it exceeds human ability and speed of image recognition. The technology can be used in a wide variety of areas - for example in parallel kinematic robots used in sorting or in complex, alternating, autonomously performed assembly steps. In summary, it can be said that machine vision can contribute to a more efficient use of cobots in SMEs - humans can be supported even better by cobots at work, and some tasks can even be carried out entirely by cobots.

The digital twin

The constantly advancing networking of machines and their components as well as a continuously growing data pool are opening the door to digital twin and Internet of Things (IoT) technologies. As technological drivers, these are leading to greater production flexibility demanded by the market. The desire to offer different variants and make production lines dynamically deployable is crystallizing among companies.

The aim is for production to move away from the conventional 'one-product production line' towards a dynamically changeable production environment. This can be achieved with the help of special simulation algorithms that regularly react to change triggers and dynamically adapt the production architectures. The simulation includes structural models of machines and their components, their mechanical and electrical functions as well as models of the production system level to ensure a flexible representation of the events on the store floor.

The author Michael Hechtel is a research assistant at the Chair of Manufacturing Automation and Production Systems (FAPS) at Friedrich-Alexander-Universität Erlangen-Nürnberg.

© FAPS

In the case of using the digital twin as a prediction tool or decision aid, the literature often refers to the use of machine learning algorithms such as 'supervised classification' and 'regression'. Another necessary point for the use of digital twins is a fully networked environment in which each machine with its sensors is part of the common platform. Such an industrial environment, together with enormously increased storage capacities, results in an immense data pool for companies. Machine learning methods - in particular supervised and unsupervised classification and clustering algorithms - can be used for analysis. One area of application for such analyses is modeling, for example, which is a critical technology of Industrial AI (IAI). It reveals the laws of the production process: the deterioration process of equipment or components, the relationship between process parameters and product quality, the coupling between the operating status of the production line and the component process. Models reflect the core production process, production capacity and competitiveness of companies. These methods enable companies to identify cost-cutting potential and optimize production through a maximum of analysis options.

Networking and data are also playing an increasingly important role for companies in the Mejoin project in the field of medical technology. The use of ICTs (Information and Communication Technologies) has increased significantly in the medical sector. Areas such as telediagnosis, teledocumentation and telemonitoring are becoming more and more established. All these methods are based on the transfer, collection and analysis of data in real time. Machine learning algorithms are used here in particular to analyze data from personal health records (PHR), personal digital medical records that are filled with personal patient health data on a voluntary basis. These enable a correlation search in data records on symptoms and behavior and enable predictive functions. The value of high-quality and well-maintained patient data is growing rapidly from such applications.

Challenges of the Mejoin project

Although there is a high level of interest in the use of AI technologies among mejoin companies, the degree of implementation to date is low. This is in line with a survey conducted by the Federal Ministry for Energy and Economic Affairs: around 7.7% of companies have experience with AI, of which 12% state that AI technology is the main component of their business model (as at 2020). The most widespread are machine learning methods for image or sound recognition and knowledge-based methods such as knowledge graphs or context recognition. Text or voice recognition have been used less frequently to date.

The situation is very similar in Japan: Surveys show that the majority of large companies, both in Germany and Japan, have already started using AI technologies. The proportion is much lower among SMEs. Those responsible often lack the necessary know-how - which Mejoin, among other things, aims to impart. The great opportunities for companies through AI such as 'greater process efficiency', 'improved sales and supply chains' and 'AI-driven target marketing' are overshadowed by certain blocking factors for SMEs. These are (in order of decreasing importance): 'a lack of know-how and skilled labor', 'a lack of high-quality data' and 'data security concerns'. In Germany in particular, where there is currently already a shortage of around 100,000 IT specialists - and the trend is rising - the lack of expertise and qualified employees plays a particularly prominent role. As the struggle on the labor market to find the right skilled workers has a negative impact on the attractiveness of AI technology, especially for SMEs, alternative ways of building up expertise play a particularly important role. Cooperation between companies at different levels of the value chain or in a cross-sectoral manner can represent such an alternative. Particularly in countries such as Germany, where more than 90% of companies are SMEs, cooperation projects such as Mejoin therefore play a major and important role in the development of new basic technologies in SMEs.

The consortium

The Mejoin project is primarily driven by the core consortium, which includes the Chair of Manufacturing Automation and Production Systems (FAPS) at Friedrich-Alexander-Universität Erlangen-Nürnberg, Saitama University, the Saitama City Foundation for Business Creation and the Cluster Mechatronic and Automation (CMA).

The core consortium of the Mejoin project comprises the Chair of Factory Automation and Production Systems (FAPS) at Friedrich-Alexander-Universität Erlangen-Nürnberg, Saitama University, the Saitama City Foundation for Business Creation and the Cluster Mechatronic and Automation (CMA).

© FAPS

The FAPS Chair is a teaching and research institute for automation technology and mechatronic systems. With over 100 employees from more than ten countries, the FAPS has specialist expertise in the fields of mechatronics, automation and AI. It has already gained valuable project experience in the organization of large-scale projects such as the Bavarian Technology Center for Electric Drive Technology (E|Drive-Center), the Green Factory Bavaria research association and the Bavarian Technology Center for resource-saving and energy-efficient living (E|Home-Center) and has created a close-knit network of qualified partners.

With Prof. Dr. Watanuki from Saitama University, an experienced partner in the technological innovation development of mechatronic systems (especially intelligent robotics systems), brain-machine interface technologies, vehicle interface design and medical technology could be won for the project network.

The Saitama City Foundation for Business Creation (SBFC) is the business development institution of Saitama Prefecture and supports small and medium-sized enterprises in innovation, management and globalization issues. It is in close contact with Saitama University.

The Cluster Mechatronics & Automation (CMA) is a platform and forum for the definition and implementation of measures that serve the progress of mechatronics and related fields. The CMA has been part of the business and innovation promoter Bayern Innovativ since January 2021 and currently brings together around 130 cluster partners.