A human-centric preference-based workforce dispatching with Machine Learning using the example of an assembly process

Aug 5, 2026

Human-centric Industry 5.0 focuses on supporting and empowering workers through value-driven production. This paper proposes a human-centered workforce dispatching approach that accounts for workers’ preferences, skills, availability, and individual strain in each task. Worker strain is predicted via machine learning to enable reduced strain, increased autonomy, and production adaptation to individual needs. Practical implementation and evaluation are carried out in a logistics learning factory.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Abstract

Introduction

Industry 5.0 and human-centricity

Stress versus strain definitions

Existing approaches and research gap

Proposed approach overview

Preference-based dispatching

Strain prediction with machine learning

Digital representation and evaluation setting

[Available online at](Available online at www.sciencedirect.com)[ www.sciencedirect.com](Available online at www.sciencedirect.com) ScienceDirect Procedia CIRP 126 (2024) 757–762 17th CIRP Conference on Intelligent Computation in Manufacturing Engineering (CIRP ICME‘23) A human-centric preference-based workforce dispatching with Machine Learning using the example of an assembly process Anja Kneissla,b, *, Günter Bitscha, Johannes L. Joosteb aReutlingen University, Alteburgstraße 150, 72762 Reutlingen, Germany bStellenbosch University, Private Bag X1, Matieland, 7602, Stellenbosch, South Africa * Corresponding author. Tel.: +49-176-43948935. E-mail address: anja.kneissl@student.reutlingen-university.de Abstract Human-centric in the context of Industry 5.0 aims to support and empower workers. This paper presents a human-centered approach to workforce dispatching. Thereby, workers’ preferences, skills, availability, and individual strain in the task are considered. To do this, the strain of the workers in the task is predicted using Machine Learning. In this way, the strain on the worker can be reduced by raising autonomy and production can be adapted to the worker’s needs. The practical implementation and evaluation are done in a logistics learning factory. © 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ([https://creativecommons.org/licenses/by-nc-nd/4.0](https://creativecommons.org/licenses/by-nc-nd/4.0)) Peer-review under responsibility of the scientific committee of the 17th CIRP Conference on Intelligent Computation in Manufacturing Engineering (CIRPICME‘23) Keywords: Industry 5.0; Human-Centric; Dispatching; Stress; Strain; Machine Learning 1. Introduction Industry 5.0, a relatively new concept developed by the European Commission, aims to shift the focus from technology-driven Industry 4.0 to be value-driven. Thereby, it focuses on three core values: human-centricity, sustainability and resilience [1] . Because humans are the most flexible part of production [2], the manufacturing sector needs to become more human-centered. Human flexibility is particularly important in assembly, as it is mostly non-routine work requiring humans to adapt to different assembly applications [3] . Even though humans play a key role in assembly, the stress level of assembly workers is higher than the stress level of workers in other positions [4] . Reducing stress of assembly workers is important for employees and companies since it hasan impact on long-term health [5], the commitment of employees to the company [6], as well as on job satisfaction [7], which impacts employee performance [8] . There are different meanings of stress in scientific, as well as non-scientific contexts. The ISO 10075-1:2017 standard defines the term mental stress as the “total of all assessable influences impinging upon a human being from external sources and affecting that person mentally”, while mental strain is defined as the “immediate effect of mental stress within the individual depending on their current condition” [9] . Per the definition, the research focused on in this paper is about strain rather than stress. There are existing approaches to reducing the strain on production and assembly workers. Most approaches aim to reduce the physical strain on workers, for example with collaborative robots [10, 11], exoskeletons [12] and the application of ergonomic design principles [13, 14] . However, there are only a few approaches focussing on reducing the mental strain on workers. These are mostly about digital assistance systems that provide information [15–17] . Providing only information to reduce strain is not enough, as strain depends on individual human factors [9] and is highly complex. 2212-8271 © 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ([https://creativecommons.org/licenses/by-nc-nd/4.0]

FAQs

What problem does the paper address in assembly work?

Assembly workers experience higher stress and strain than many other roles. The paper targets reducing workers’ strain by adapting task assignments to individual needs.

How does the approach incorporate human factors into workforce dispatching?

The method considers workers’ preferences, skills, availability, and individual strain impact for tasks, aiming to increase autonomy and control.

How is machine learning used in the proposed system?

Machine learning predicts a worker’s strain for a given task, enabling selection of workers with the least expected strain and supporting strain-aware dispatching.

Aug 5, 2026

Upload by Clementine