Designing the interaction with Intelligent Decision Support Systems in Control Rooms: Challenges, Strategies, and Insights for Railway Applications

Control rooms are critical environments for monitoring and managing complex socio-technical systems across industries such as transportation, energy, and public safety. Decision Support Systems (DSS) play a pivotal role in assisting operators by processing vast amounts of data, streamlining decision-making processes, and reducing response times. The integration of AI into DSS, creating Intelligent DSS, introduces new challenges, particularly regarding explainability and trustworthiness. Operators must not only interpret complex AI-driven recommendations but also rely on them in high-stakes, time-critical scenarios. For instance, intelligent alarm management systems in railway control rooms are designed to help operators prioritize, filter, and manage alarm floods, reducing cognitive overload. However, their effectiveness heavily depends on aligning with operators' cognitive needs, maintaining situational awareness, and fostering trust in automated recommendations. This context presents new significant challenges for designing effective interactions between control room systems and operators, that differently for the back-end AI-based DSS solutions, remain less clearly defined. This gap complicates the development of clear strategies for ergonomic interaction design and their subsequent assessment. This study addresses these challenges through a systematic literature review, focusing on works within the domains of human factors and ergonomics. The review explores the following research questions: What are the main ergonomic issues identified in the current state of the art regarding operator interaction with Intelligent DSS in control rooms? What are the key interaction strategies proposed to address these issues, and what performance indicators have been identified? Performance indicators are defined operationally and accompanied by detailed methodologies for their calculation, ensuring their applicability to other design projects. These indicators include measures encompassing both objective and subjective aspects, related to situational awareness metrics, and trust in AI systems. By synthesizing research perspectives and providing actionable guidelines, this study offers a foundational reference for ergonomic design efforts in control room environments. It seeks to overcome current limitations and advance the development of safer, more efficient, and operator-friendly systems, with a particular focus on railway applications.