Human-centered AI applications in production

Artificial intelligence (AI) is finding its way into industrial production. There, it is performing more and more tasks that would normally require human intelligence. AI applications can be found along the entire value stream - from logistics to manufacturing and assembly processes. Their potential to optimize processes and business models appears to be huge. They can improve the productivity and efficiency of industrial production processes and contribute to tasks that cannot be performed economically by purely human labor.

Nevertheless, many production companies are reluctant to introduce AI applications. They do not know how to start AI projects and how to approach them systematically. After all, data-intensive optimization approaches in the production sector are initially unknown. In addition, many companies are in the middle of digitalization activities and have not yet achieved short-term business success. This feeds their skepticism about tackling new, seemingly highly complex tasks associated with the introduction of AI applications. Nevertheless, manufacturing companies should consider the potential opportunities offered by AI applications at an early stage and clarify their areas of application.

If you want to realize future potential, you need to create the right conditions now. In addition to the issue of a systematic introduction process, companies must ensure that employees are involved in and actively support AI initiatives. Only in this way can AI applications in operational use realize the intended potential benefits. In addition to technical issues, companies are therefore required to design AI applications, processes and forms of interaction in a human-centered way.

It is true that every company will have to find its own way of introducing AI applications. Nevertheless, the study report produced as part of the project offers a generic orientation for the beneficial introduction of human-centered AI applications into industrial production processes. A pragmatic action guide shows how technical functions and human potential can be combined for the benefit of companies, employees and customers. It focuses on the following topics:

• Identification of AI applications in production
• Multi-dimensional evaluation of AI applications
• Selection of suitable processes for AI applications
• Data selection, data pre-processing and data processing
• Human-centered design of work processes and human-machine systems
• Impact monitoring
• Implementation strategies and change management

The guidelines are supplemented by literature references and pragmatic templates. It is based on an intensive exchange of experience with company practitioners in workshops, company projects and expert interviews.