Beyond Automation : Guidelines for a Human-Centered Multi-Agent System for Coordinated Decision Making under Industrial Settings

Sammanfattning

This thesis investigates how Multi-Agent Systems (MAS) can be designed in a human-centered manner to support coordinated decision-making in complex industrial environments, such as fleet management, logistics, and manufacturing. The primary research question guiding this investigation is: How can a multi-agent system be designed to be human-centered and support coordinated decision-making within industrial contexts? With the increasing complexity brought by Industry 4.0 technologies, including cyber-physical systems, IoT, and real-time analytics, there is a clear necessity for advanced decision-support systems that enhance human decision-making capabilities and MAS in particular have demonstrated potential in this regard. However, existing MAS implementations often neglect human-centered factors, leading to issues such as mistrust, cognitive overload, and suboptimal interactions between humans and agents. To address this gap, this research employs qualitative methodologies grounded in Human-Centered Design (HCD) and Value Sensitive Design (VSD), engaging stakeholders directly through contextual inquiries and in-depth interviews within fleet operation support services. Findings reveal critical factors essential for a human-centered MAS, including transparency, iterative communication, behavioral alignment, context-aware assistance, multimodal interaction, and ethical design practices. The thesis provides comprehensive guidelines addressing these factors and highlights their support in achieving effective coordinated decision making. By doing so, this work contributes to Human-Computer Interaction (HCI) and Artificial Intelligence (AI) fields, more specifically, the subfield of Human-Centered AI (HCAI) by offering guidelines for developing MAS that integrate seamlessly into human workflows, enhancing both efficiency and trust in industrial decision-making contexts.