Artificial intelligence in mechanical engineering

Mechanical enterprises are the main users of artificial intelligence technology, and also the suppliers of industrial solutions. They are also the core link to promote and apply artificial intelligence in the industrial value chain. In the form of embedded artificial intelligence, machines and equipment bring artificial intelligence technology to users and bring them to all walks of life. Mechanical engineering has accumulated experience in the efficient integration and responsible design of human-machine collaboration such as robotics, automation, and sensor technology. The German Federation of Machinery and Equipment Manufacturing must support the application of member companies, but in the political and social environment, it is also important to expand the scope of artificial intelligence technology acceptance and form a social system.

Mechanical Engineering: Amplification Effects of Users and Artificial Intelligence Technology

For mechanical engineering, artificial intelligence is first and foremost an opportunity to stay ahead of the world. Artificial intelligence helps to improve work efficiency and develop new business models. The intelligent functions of embedded artificial intelligence solutions can also optimize production processes and expand the use and service range of machines. Artificial intelligence will determine the impact of future products and processes in mechanical engineering. Therefore, the foundation for building superior artificial intelligence relies on both existing technical expertise and application expertise, and mechanical engineering is the basis for this, and will play a pivotal role in the application of cross-industry and cross-sector artificial intelligence. This is not only good for mechanical engineering companies and their customers, but also has great potential for saving materials and energy, improving decision-making, controlling resource shortages and climate change. On the other hand, if the opportunities brought about by artificial intelligence cannot be successfully used, the leading position of European engineering companies will certainly decline and lose to competitors in other parts of the world. Therefore, companies, mechanical research institutions and policy makers must integrate artificial intelligence into mechanical engineering.

Mechanical engineering must also be brave enough to take responsibility for introducing new technologies – whether it is to ensure mechanical safety or to communicate with society. However, the German Federation of Machinery and Equipment Manufacturers believes that artificial intelligence is not a new independent policy, but a key technology with horizontal significance. Not only digital topics such as data management, digital platforms, network security, and IT infrastructure should be concerned with it, but also the “traditional” industry activities such as product safety, mechanical safety, operating environment design, and standardization. For example, the legal requirements for product safety and EU Harmonized Regulations already cover mechanical artificial intelligence applications.

In order to consider comprehensively, make full use of opportunities, and objectively analyze risks, social dialogue must be based on facts based on the participation of mechanical engineering and industry users.

The concept and application of “artificial intelligence”

If “artificial intelligence” refers to a humanoid system with unlimited autonomy, then a practical digital policy should not be dominated by it. It is important to discuss the conceptual differences between general artificial intelligence and narrow artificial intelligence by seeking truth from facts: general artificial intelligence refers to attempts to simulate humanoid intelligence—the ability to plan, make decisions, etc. in an uncertain state or in pursuit of complex goals, but When it can be realized, or whether it is possible to achieve it, there are still many different opinions. There is no final conclusion. On the other hand, in the specific application process, narrowly defined artificial intelligence, such as speech recognition, pattern recognition, error analysis, etc., is now being developed. Moreover, the law stipulates that such “artificial intelligence” is limited to the intended use and specifies the specific requirements for developers, while physical processes, operational requirements and technical standards also explicitly limit the use of artificial intelligence for industrial processes. Machine manufacturers are keen to control all the functions of the machine, especially the functions that artificial intelligence generates or changes. Therefore, the actual policy discussion can only be based on the current narrow form of artificial intelligence – there is a certain degree of autonomy in specific applications, but there is no human intelligence – as the basis, “artificial intelligence” in this article refers to narrow artificial intelligence.

“Machine learning” is a form of narrowly defined artificial intelligence. It has been implemented and can be evaluated objectively. Machine learning is based on statistical algorithms that allow software applications to learn independently based on pattern recognition. Current industrial and mechanical engineering has been using machine learning to solve specific technical or economic problems. For example, in the field of quality assurance, image processing methods have been used to check the surface condition by machine learning, and the image processing efficiency has a great potential for improvement. Another example is process optimization for complex machines: machine learning based on sensor data can provide valuable information, reduce debugging time, and discover unknown sources of error. Predictive maintenance is a data assessment that aims to improve the efficiency of operations, maintenance, and repair processes. Its successful application of algorithms has become a model for the industry. Key indicators, by evaluating data such as ERP, can help optimize internal production structures and processes, such as providing data during the use phase, providing information for innovation and process improvement to help improve product development and management. In sales and planning, artificial intelligence tools intelligently configure machines to deliver significant business value potential.

These examples show that artificial intelligence is used in industrial applications and has many opportunities that are expected to bring considerable benefits. However, this also shows that ethical discussions are absolutely necessary in many cases, but the discussion of each application situation is not exactly the same; especially in industrial applications, such problems are generally less important. Therefore, the ethical discussion about artificial intelligence cannot be rushed across the board, otherwise it will prematurely limit or unnecessarily limit the scope of innovation of artificial intelligence applications, and cannot quickly use artificial intelligence for promising applications.

Political field of action and core information

Industrial policy and economic policy: industry uses artificial intelligence to defend leadership

To succeed in international competition, Europe can only be competitive if it is united. The EU internal market is a model of EU success and will play a central role in this regard. Only through the cross-border plan to make the market coordinated, can we achieve the necessary scale effect and form an investment environment. National measures and even laws must be avoided. The research plan must also be universal, coordinate and communicate, and avoid duplication. In addition, bringing together outstanding talents in the scientific and industrial fields is also an important measure. Therefore, the German Federation of Machinery and Equipment Manufacturing should fully support the relevant artificial intelligence plan of the European Commission.

On the other hand, it is also necessary to focus on the world and pay attention to international competition. Europe must face international challenges and cannot be limited to defensive strategies. The solution consists of developing a vision of industrial policy based on the strengths of Europe, using unique expertise and industrial capabilities. Although competitors in the B2C sector may be ahead of Europe, industrial and mechanical engineering can be a pioneer in this field as long as the route is set correctly. The European Commission’s “all-European measures” for artificial intelligence is the right path. However, the importance of “Europe as a whole” cannot be separated from the international market. It is necessary to consider the role that international standards can play and avoid the formation of hostile innovation atmosphere. It is necessary to continue to operate international institutions and platforms that explore artificial intelligence and necessary guidelines. It is also necessary to formulate policies such as “equitable environment” for international digital competition and digitalization of trade.

Research and innovation: unlocking the power of innovation

In order to take advantage of Europe and the potential of artificial intelligence in terms of competitiveness and efficiency, artificial intelligence research must be both horizontal and fundamental, as well as specific applications for business and industry. Not only must the algorithm be developed, but also the specific problem. Acquire and select data on a case-by-case basis and guarantee data quality. When artificial intelligence is used in a highly developed industrial technology system, its safety, process reliability, and quality must also meet high standards.

Priority should be given to applications and industries that are expected to spread rapidly and effectively scale up—for example, artificial intelligence is used in product development design or new business models in industrial production. Artificial intelligence is an interdisciplinary field that requires data experts or artificial intelligence experts to work closely with people in other disciplines. Therefore, it should not be restricted by the state or discipline, but should give priority to the way to promote cooperation. The EU should also play its own advantages in the study of artificial intelligence and carry out cross-border cooperation. But traditional research funding alone does not generate enough power and a wide range of artificial intelligence applications. Therefore, it is very important to play the flexible innovation ability and creativity of small enterprises. It is necessary to formulate relevant strategies, remove restrictions, release vitality, and provide tools for the dynamic development of start-ups and innovative SMEs.

Promote the transfer to industry

Only by expanding the scope of artificial intelligence and transferring it to a large number of small and medium-sized enterprises and medium-sized industrial enterprises can the artificial intelligence be successful in Europe. It is therefore important to ensure the efficiency of technology transfer and to lower the barriers to access to technology, projects, outcomes and networks. One of the basic measures is to test the process model and the business model with test centers and competence centers (such as the “Digital Innovation Center”) – as long as they are in an industrial environment, providing a precise form of practice-oriented. It is also important to have the best practices and efficient tools to enable companies to get the knowledge they need to solve the shortage of available information technology experts. “Artificial Intelligence Self-Service” or “Guided Analysis” helps make it easier for business experts to apply artificial intelligence. Companies can work with AI experts (possibly from outside the enterprise, specific technology transfer areas, and professional training) to define requirements and evaluate solutions. The state of development of the cutting-edge technology in the standard, describing the possible application forms, is also important for large-area transfer of technology.

Employment situation: artificial intelligence needs people

The discussion of artificial intelligence has once again fueled concerns about the replacement of people by machines. However, new business opportunities will emerge, productivity will increase, Europe’s production bases and technology bases will not be shaken, and new jobs will emerge. The German Federation of Machinery and Equipment Manufacturing believes that some traditional positions will disappear and other new jobs will emerge. However, the impact ratio of the two is currently difficult to predict. The results of a large number of studies are different. Some believe that the newly created posts will fully complement the lost posts, and some believe that the overall position will be completely changed. However, looking back at history, it can be proved that the positive view is reasonable: despite the high rate of application of robots in Germany, the number of employed people has never been higher. How the employment situation develops depends on many factors that are not yet clear, but one of the factors that can be affirmed is that artificial intelligence is being studied in various industrial zones around the world. If you do not participate in research and development, jobs lost in Europe will be Transfer to other countries.

From the perspective of technical practice, artificial intelligence technology is bound to be inoperable without the participation of others. Artificial intelligence is a powerful and efficient tool, but it’s still just a tool. Artificial intelligence analysis only provides predictions and possibilities, and cannot make decisions similar to humans. Humans still have to assess conflicting goals, measure various factors, and ultimately make major decisions and take responsibility. As data processing is intelligent, information technology provides a broader space for enhancing human creativity. Although human beings are irreplaceable, the content and requirements of the work will change. The need to coordinate with the artificial intelligence system to coordinate communication, efficient communication, and self-determination of responsibility has never been improved. The harmonious cooperation between humans and systems depends to a large extent on whether the legal, employment, and educational policy framework conditions will adapt to future needs and adjustments. The existing frameworks in some areas need to be appropriately revised. In particular, modern labor market policies need to be revised. Adapt to the future market.

The more emotional view is that artificial intelligence will soon eliminate humans and even dominate the labor force without a technical foundation. The correct policy measure is to monitor the actual development of technology and society, for which policy makers should make further decisions based on a calm analysis. Under no circumstances should legal policies hamper or hinder artificial intelligence applications. Horizontal regulations such as “robot tax” or surface measures are not the solution. Legislators should be in a complementary position in the field of artificial intelligence applications. If direct participants cannot find a solution, legal policies can directly intervene in the market, and the development and design space at the enterprise and individual levels should be emphasized.

Master plan for education and training

Even without massive unemployment, the labor market seems to have changed radically. Artificial intelligence and automation will change the task of work and need to learn new skills and knowledge. This not only promotes the research and education of cutting-edge technologies, but also trains information technology experts. It also needs to invest heavily in education and training to provide low-level applied courses. In the context of artificial intelligence, its own education and training becomes more important – because the application of machine learning (such as “supervised learning” or “enhanced learning”) does not replace humans, but requires human beings to become trainers or managers: Artificial intelligence systems are complex analytical systems that require developers and operators to have the appropriate skills. Therefore, not only information technology experts and programmers must have artificial intelligence knowledge, the application of artificial intelligence at all levels, functions, and staff of all departments must have the ability to work artificially, which requires the development of a digital planning qualification to meet the overall planning challenges.

Avoid putting together digital policies and data policies

Artificial intelligence is part of digitalization and requires the provision of appropriate system conditions for network connectivity and data exchange, such as sharing machine data. One of the core prerequisites for the successful application of artificial intelligence is to increase the trust between business partners and the reliability of the political framework.

Artificial intelligence requires a holistic view of digital policy to avoid the contradiction, obstacles and legal ambiguity caused by isolated measures. It is clear that existing laws or planned laws are interfering with the application of artificial intelligence in ways that are not necessarily successful (General Data Protection Regulations, Platforms, Cybersecurity, Copyright, Electronic Privacy). Data and data exchange are especially important for developing artificial intelligence applications. Only transnational measures across Europe are sustainable, so the German Federation of Machinery and Equipment Manufacturers supports the EU’s European-wide, freely liberalized data field. If data can be shared as widely as possible and not hidden in isolated or proprietary models, artificial intelligence can bring more benefits to the macro economy. The German Federation of Machinery and Equipment Manufacturers supports the promotion and enhancement of data exchange – for example, open data measures in publicly funded research (“as open as possible, on-demand confidentiality”), or through data governance models (with the interaction of technology and contract terms) Basic), both to improve the application of data and to improve the protection of investment and intellectual property.

However, in principle, legislators should avoid rashly intervening in the development of the data market. Especially in the B2B field, the diversification and dynamic development of business models and applications requires the greatest possible flexibility and legal certainty. The principle of freedom of contract should be further strengthened, and any rules that restrict freedom should be reviewed. Only when the market failure or concentration trend leads to unfair negotiations, it is necessary to review the fairness of the contract application conditions.

Regulatory framework: starting with freedom, not starting with restrictions

In order for people to accept and successfully apply artificial intelligence technology, human-machine cooperation must be realized. “People-oriented artificial intelligence” may be a method, but it may mislead people into thinking that they need to protect humans from artificial intelligence. But man-machines and progress are closely linked: people make machines and machines to improve their lives or simplify their lives. In this sense, artificial intelligence is just an improvement of the machine. Therefore, the German Federation of Machinery and Equipment Manufacturing supports in principle the idea of ​​seeking opportunities and objectively analyzing risks, but does not support abuse and misapplication – as is the case with other technologies. Artificial intelligence is at risk of opacity, discrimination and manipulation, and it is necessary to discuss its transparency and comprehensibility.

Share