AMTFuP – Analyse mensch- und technikbezogener Fehlerursachen in der Produktion

Motivation

The human will continue to incorporate a critical role in interconnected production systems. Within the framework of Industry 5.0, there is a heightened emphasis on the role of individuals in cyber-physical production systems.

As performance decreases with advancing age, increasing strain and sickness rates of employees in manual assembly are expected in the long term due to the demographic change. More complex products and pressure regarding quality, costs, and time pose further challenges for companies and requirements on employees.

Due to the declining performance capacity of the workforce coupled with increasing requirements, a long-term increase in the number of human errors can be expected.

Objectives

The aim of the project is to reduce the occurrence of human errors in production. To achieve this goal, it is necessary to be able to predict human errors. Therefore, an AI model must be developed to predict human errors in manual assembly. With the help of this model, errors can be predicted based on error-relevant factors of the production system and the human. In addition, suitable measures need to be developed to prevent predicted errors. The countermeasures are introduced in case of an increased risk of human errors to counteract their occurrence.

Approach

Within the framework of the AMTFuP project, the aim is to investigate the causes of human errors in production and to elaborate error-relevant factors of the production system and the human. This is accomplished through a detailed analysis of error research and a human-centred consideration of the workplace and the working environment. The error-relevant factors must be prioritised systematically to enable an individual and effective solution depending on the application. With the help of knowledge of errors and a methodology for selecting suitable tools to collect error-relevant data – such as stress and strain – intelligent devices can be used to collect data on both the production system and the employees. An AI-based prediction model will process this data to predict human errors through pattern recognition.

Countermeasures will be developed and validated based on the causes of the errors. Combined with the prediction model, a reduction of human errors is achieved.