Concept of a data-based approach for the prediction and reduction of human errors in manual assembly

The manufacturing industry faces numerous challenges, with societal and economic developments affecting the entire industry. On the one hand, rising pressure regarding costs, quality, and time due to international competition and increasing product variety, as well as individualization, is increasing the requirements regarding the performance level of assembly workers. On the other hand, some industrialized countries like Germany face a shortage of skilled workers and a demographic change. The resulting discrepancy between performance requirements and the actual capability of workers can induce increasing numbers of human errors, e.g., due to stress reactions, resulting in rework and scrappage. Hence, this article presents a concept for reducing human errors in manual assembly. Firstly, the systematic identification and prioritization of factors causing human errors are described. Next, the collection of data, e.g., by using intelligent devices, on error-causing factors - like task load or mental strain - is presented. After that, a data-based approach to predict human errors using an AI model is outlined. The systematic derivation of countermeasures is recommended to reduce the occurrence of human errors. The methodology aims to increase the profitability of companies by lowering scrappage and rework through error reduction.