2025 IEEE Conference on Cognitive and Computational Aspects
of Situation Management (CogSIMA)

June 2-5, 2025 | Duisburg, Germany

Photo: Ilja Höpping, Stadt Duisburg

Tentative Program

tbc

Berlin time Monday, June 2 Tuesday, June 3 Wednesday, June 4 Thursday, June 5
8:45 ‑ 9:00   Conference Opening    
9:00 ‑ 10:00 T2: Tutorial 2
T1: Tutorial 1
Keynote 1 Keynote 2 Keynote 3
10:00 ‑ 10:20 Coffee Break Coffee Break Coffee Break
10:20 ‑ 11:50 S1: Information Fusion and Situation Management S3: Situation Awareness and Cognitive AI S6: Cognitive Modeling and Decision Making
11:50 ‑ 12:30 Lunch Break Lunch Break Lunch Break
12:30 ‑ 13:00 Lunch Break
13:00 ‑ 14:00 S2: Special Session S4: Human-AI and Human-Robot Teaming P3: Panel on Semantic Forensics
14:00 ‑ 14:30 T4: Tutorial 4
T3: Tutorial 3
Conference Closing & CogSIMA 2026 Outlook
14:30 ‑ 14:45 Coffee Break  
14:45 ‑ 14:50 S5: Poster Presentations  
14:50 ‑ 15:15 Coffee Break  
15:15 ‑ 16:15 P1: Panel on Situation Control  
16:15 ‑ 16:30 Coffee Break  
16:30 ‑ 17:15 P2: Panel on Human-AI Teaming  
17:15 ‑ 17:30    
17:30 ‑ 18:00 Icebreaker Cocktail TO1: Lab Tour Autonomous/Remote Controlled Vessels  
18:00 ‑ 19:00    
19:00 ‑ 19:30 Conference Banquet  
19:30 ‑ 20:00    
20:00 ‑ 21:00      

Monday, June 2

Monday, June 2 9:00 - 12:30 (Europe/Berlin)

T1: Tutorial 1

Detecting, Attributing, and Characterizing Generated and Manipulated Media Using Semantic Forensics
Michael Kozak, Lockheed Martin Advanced Technology Laboratories, USA

https://edas.info/web/cogsima2025/tutorial_program.html#Kozak

Monday, June 2 9:00 - 12:30 (Europe/Berlin)

T2: Tutorial 2

Decision Making Under Risk, Uncertainty, and Ignorance
Galina Rogova, Ph.D., The State University of New York at Buffalo, USA

https://edas.info/web/cogsima2025/tutorial_program.html#Rogova

Monday, June 2 12:30 - 14:00 (Europe/Berlin)

Lunch Break

Monday, June 2 14:00 - 17:30 (Europe/Berlin)

T3: Tutorial 3

Cognitive Systems Design
Prof. Antonio Lieto University of Salerno, Italy

https://edas.info/web/cogsima2025/tutorial_program.html#Lieto

Monday, June 2 14:00 - 17:30 (Europe/Berlin)

T4: Tutorial 4

Neuroergonomics for Situation Awareness: Applying Wearable Neurotechnologies in Complex, Real-world Environments
Hasan Ayaz , Drexel University, Philadelphia, PA, USA

https://edas.info/web/cogsima2025/tutorial_program.html#Ayaz

Monday, June 2 17:30 - 19:30 (Europe/Berlin)

Icebreaker Cocktail

Tuesday, June 3

Tuesday, June 3 8:45 - 9:00 (Europe/Berlin)

Conference Opening

Tuesday, June 3 9:00 - 10:00 (Europe/Berlin)

Keynote 1

Tuesday, June 3 10:00 - 10:20 (Europe/Berlin)

Coffee Break

Tuesday, June 3 10:20 - 11:50 (Europe/Berlin)

S1: Information Fusion and Situation Management

10:20 Enhanced Multi-Source Information Fusion for Advanced Air Mobility Through AI-Based Situation Assessment and Incorporating OSINT Data
Christoph Allig (Hensoldt Sensors GmbH, Germany); Ulrich Lode (HENSOLDT Sensors GmbH, Germany); Christoph Schmid (Hensoldt Sensors GmbH, Germany)

We present an overview of the Enhanced Multi-Source Information Fusion (EMSIF) project, as well as the research work carried out during the project. The aim is to create an improved situation picture for the future application domain Advanced Air Mobility. The improvement comprises two parts. On the one hand, new types of data sources are integrated, such as OSINT data and Background Knowledge (BK). On the other hand, conventional Multi-Source Information Fusion (MSIF) is often limited to JDL Level 1, which estimates the states of individual entities. Conversely, Level 2 considers the relations between the individual entities to detect critical situations and draw conclusions. That allows the operator to detect and evaluate hazardous situations more easily, even in crowded environments.

10:40 Towards Cognitive Situational Awareness in Maritime Traffic Using Federated Evidential Learning
Shang Gao and Zhixin Huang (University of Kassel, Germany); Ghassan Al-Falouji (Kiel University & Intelligent Systems Research Group (INS AG), Germany); Bernhard Sick (University of Kassel, Germany); Sven Tomforde (University of Kiel, Germany)

The increasing complexity of maritime navigation and the shift towards (semi)autonomous systems necessitate enhanced situational awareness (SA) to ensure maritime safety. This transition introduces new requirements for situation modelling and SA, particularly in busy waterways. To address these challenges, we present the Federated Evidential Learning for Anomaly Detection of Ship Trajectories (FEAST) framework, which integrates Federated Learning and Evidential Learning to provide a privacy-preserving, collaborative, and uncertainty-aware approach to out-of-distribution (OOD) anomaly detection in maritime traffic. FEAST utilises data from the Automatic Identification System from the Kiel region, Germany, which exhibits unique characteristics of dynamic and heterogeneous maritime activity due to its connection with the traffic-dense Kiel Canal. Our extensive evaluations demonstrate that FEAST improves OOD anomaly detection by leveraging epistemic and aleatoric uncertainty estimates, outperforming baseline methods such as Denoise AutoEncoders and Variational AutoEncoders. Consequently, FEAST forms a solution to reliable and interpretable maritime traffic anomaly detection, supporting enhanced SA in maritime operations.

11:00 Enhanced Shared Situational Awareness and Decision Support in Maritime Firefighting: Insights from the Overheat Project
Hasan Ahmad (Ecole National Superior Maritime, France); Pedro Merino Laso (French Maritime Academy, France)

Shared Situational Awareness (SSA) plays a critical role in effective fire management on container ships. This paper examines the challenges of maritime fire incidents, emphasizing the prevalence of human error and the limitations of traditional fire detection and coordination methods. It stresses the importance of Endsley's three-level model of Situational Awareness (SA) in maritime fire scenarios and proposes a model for SSA development specific to such emergencies. The Overheat Project is introduced, highlighting the development of a Digital Solution (DS) to address enhanced SSA through real-time data integration and a collaborative platform. This platform connects vessels and on-shore data systems, providing a common operational picture (COP) for all stakeholders involved in fire response. The architecture of the DS is described, outlining the onboard and ground-based components that facilitate seamless data exchange. The paper concludes by highlighting the potential benefits of this solution in improving fire safety and response capabilities on container ships.

11:20 Trust in Fusion-Driven Human-Machine Environments
Galina L. Rogova (University at Buffalo, USA)

Trust can be defined as an information quality characteristic representing a subjective level of belief of a user (either human or automatic) that the information he is using can be admitted into the system, transferred between system processes, or used for making decisions. The trustworthiness of an automatic agent is defined by the level of reliability considered in a particular context and based on domain knowledge, and statistical information obtained from previous experience/experiments. The problem of defining trust in information provided by humans is more difficult since their characteristics can be unknown, information can be manipulated or affected by multiple biases. The paper discusses the problems of trust representation, incorporating it into a fusion-based system, and introduces an approach to modeling trust and distrust based on a unified framework of the Transferable Belief modal and Belief-based argumentation.

Tuesday, June 3 11:50 - 13:00 (Europe/Berlin)

Lunch Break

Tuesday, June 3 13:00 - 14:50 (Europe/Berlin)

S2: Special Session

Enhancing Situational Awareness, Safety, and Trust through User-Centered Design in Collaborative and Intelligent Systems
13:00 Understanding the Factors Supporting Eco-Driving Decision-Making: How Technology, an Ecological Mindset, and Cognitive Flexibility Promote Sustainable Driving
Eva Gößwein (University of Duisburg-Essen, Germany); Julia Braun (Brandenburg University of Technology Cottbus-Senftenberg, Germany); Jana Thin and Magnus Liebherr (University of Duisburg-Essen, Germany)

As sustainable motorized individual transport becomes essential for achieving CO2 reduction targets, this study investigates the acceptance of an eco-driving app that will maybe provide a practical solution to improve sustainable diving. The app, building on currently available technologies, was examined using an integrated model of the Technology Acceptance Model (TAM) and Trust in Technology. The study assesses the technology's acceptance and influence on future eco-driving intentions. In addition to TAM's established variables - Perceived Ease of Use and Perceived Usefulness - Trust in Technology emerged as a significant predictor of app acceptance. The study further explored the roles of Environmental Awareness and Cognitive Flexibility, two factors previously unexamined in eco-driving research, as motivators of sustainable driving intentions. Findings indicate that app acceptance, alongside Environmental Awareness and Cognitive Flexibility, supports the Intention to Perform Eco-Driving. Practical implications emphasize the importance of a design that fosters user trust, while theoretical implications suggest that future research should consider cognitive and mindset factors in addition to technology acceptance when investigating eco-driving intentions. These results underscore the relevance of our study for both practitioners and researchers in advancing sustainable motorized individual transport.

13:20 Empowering Trust: the Role of Adaptable Design in AI Systems
Verena Staab (Duisburg Essen University, Germany); Ilka Hein (Ludwig-Maximilians-Universität München, Germany); Maike Ramrath (University of Wuppertal, Germany); Lea Schlüter, Alina Stuckstätte and Maximilian Hohn (University of Duisburg-Essen, Germany); Philipp Sieberg (Schotte Automotive GmbH & Co. KG, Germany); Magnus Liebherr (University of Duisburg-Essen, Germany)

The integration of AI in workplaces has increased automation but often at the cost of transparency, potentially undermining user trust. Adaptable, user-centered systems address this challenge by enhancing situation awareness and tailoring interactions to user needs through adaptable elements that allow users to customize system settings. Unlike most XAI research, which emphasizes technical transparency, these systems prioritize the user experience to enhance understanding and trust. In an online study with a between-subjects design, 197 participants interacted with an adaptable or non-adaptable system for decision making to examine the influence of its features on perception (transparency, fairness, control), trust, and intention to use. Results indicate that adaptability positively affects perceived transparency, fairness, and especially control. Furthermore, intention to use was positively influenced by the mediators of perceived fairness, control, and the serial mediator trust. This study underscores the importance of adaptable design elements in human-computer interaction, demonstrating that they enhance user perception and intention to use, whereas AI systems that restrict user involvement and autonomy risk diminishing both trust and intention to use.

13:40 Designing Human-Robot Collaboration for Semi-Automated Liquid Cargo Handling
Markus Nieradzik (University of Duisburg-Essen, Germany); Verena Staab (Duisburg Essen University, Germany); Nils Nover (University of Duisburg-Essen, Germany); Tobias Bruckmann (Universität Duisburg-Essen, Germany); Dieter Schramm (University of Duisberg-Essen, Germany)

Today, the handling of liquid cargo in inland navigation, especially the mounting and unmounting of the loading hoses, is a predominantly manual and physically demanding operation that carries a high risk of damage due to human error. In response to these problems, the ongoing research project "CoboTank" focuses on the development of a collaborative robot application and the application-specific robot system for this use case. In particular, the implementation of the collaborative robot aims to reduce physical workload on workers while simultaneously minimizing the number of personnel required for the task, without increasing the risk of damage. The implementation of a collaborative robot application requires the formulation of a robust safety concept for human-robot collaboration (HRC) and the systematic allocation of tasks between the human operator and the robot. This contribution addresses these two key aspects. The developed safety concept uses the safety mode of hand guided controls according to standard ISO 10218 together with a user-centered human-machine interface, designed especially for this application. The task allocation approach first considers feasibility and economic efficiency in the evaluation between manual, automated or HRC-based realization of a work step and afterwards implements the HRC-based work steps based on human factors. Both presented solutions not only aim to achieve physical relief and ensure user-friendliness but also play a pivotal role in enabling the remaining worker to maintain sufficient situation awareness to make appropriate decisions in emergencies and unforeseen situations. This enables the prevention of personal injury, as well as environmental or material damage. To ensure practical applicability, the solutions developed were validated by inland waterway boatmen trainees, as well as tank farm operators and other stakeholders in the industry. This ensures user-centered development and guarantees the transferability of the solutions to comparable applications.

14:00 Analysis of Driver Workload on Risk Perception in Non-Safety-Critical Situations with Partial Automation
Khazar Dargahi Nobari (TU Dortmund University, Germany); Torsten Bertram (Technische Universität Dortmund, Germany)

The acceptance of automated driving depends largely on the user's perception of risk and trust in the automated vehicle. One of the challenges in the design of user interfaces and interaction with the human driver is therefore to promote trust and reduce risk perception in non-safety-critical situations. The aim of this contribution is to investigate the effect of driver workload on perceived risk of drivers during partially automated driving and to attenuate the perceived risk associated with non-safety-critical situations by adapting driver-related factors. To address this objective, a section of the manD 1.0 (human driver monitoring) dataset is utilized, which includes the behavior of 39 drivers in car-following scenarios, where the perceived risk and trust in the automated system are affected by assigning a secondary auditory task to the participants. The results of the analysis show that driver interventions in automated driving, such as braking, steering, or taking control, as indicators of perceived risk, occur more frequently when drivers are not multitasking, even when the driving situation is not critical. The consideration of age, gender, and driving experience shows that driver characteristics have no significant influence on the takeover decision in this context. The proposed interaction can be integrated into a feedback loop to enhance trust when falsely high perceived risk is detected. Such an approach can be applied to a range of automation levels, including partially or entirely passive drivers, as well as to passengers in automated shuttles where the absence of a human driver may lead to increased perceived risk. Further studies are required to learn more about the qualitative and quantitative relationships between the different mental states of drivers and their improvement methods.

Tuesday, June 3 14:50 - 15:15 (Europe/Berlin)

Coffee Break

Tuesday, June 3 15:15 - 17:15 (Europe/Berlin)

P1: Panel on Situation Control

Cognitive and Computational Issues of Situation Control: 16 Years Later

Tuesday, June 3 17:30 - 20:00 (Europe/Berlin)

TO1: Lab Tour Autonomous/Remote Controlled Vessels

Wednesday, June 4

Wednesday, June 4 9:00 - 10:00 (Europe/Berlin)

Keynote 2

Wednesday, June 4 10:00 - 10:20 (Europe/Berlin)

Coffee Break

Wednesday, June 4 10:20 - 11:50 (Europe/Berlin)

S3: Situation Awareness and Cognitive AI

10:20 AXONS-3: an XAI-Augmented Approach for Advancing Trust and Transparency in 3D Brain Tumor Segmentation
Jacqueline Abyasa and Rissa Rahmania (Bina Nusantara University, Indonesia)

Early brain tumor detection remains a critical challenge in medicine due to its impact on patient outcomes. While magnetic resonance imaging (MRI) is a key tool, the challenges accumulated from the grayscale nature of MRI and high volume of data, coupled with human cognitive limitations and time pressures in radiology, create a potentially large margin of diagnostic uncertainty and error. Despite their performance, deep learning solutions face resistance in clinical adoption due to the lack of trust that roots from the opacity of such models. This research introduces the AXONS-3 workflow with the aim of bridging model outputs with the practical needs of clinicians by integrating interpretability and transparency into artificial intelligence (AI) systems for clinical decision-making. First, a 3D U-Net model is trained on the BraTS2020 dataset using T1-Gd, T2, and FLAIR MRI sequences to segment brain tumors into sub-regions of NCR/NET, ED, and ET. Then, post-hoc visual Explainable AI (XAI) techniques, including gradient-based methods and uncertainty quantification, are augmented to the workflow to interpret the process of reaching the predicted segmentation. The proposed AXONS-3 workflow provides visually intuitive feedback and justifications to foster greater stakeholder comprehension and trust, contributing to the transparency of AI-driven systems needed for reliable adoption in clinical settings.

10:40 Context-Aware Worker Assistance System in Augmented Reality Using Semantically Zoomable Digital Twin
Snehal Walunj (RPTU Kaiserslautern, Germany); Parsha Pahlevannejad, Ali Karnoub and Christiane Plociennik (German Research Center for Artificial Intelligence, Germany); Martin Ruskowski (Deutsches Forschungszentrum für Künstliche Intelligenz GmbH, Germany)

Manual maintenance and repair tasks are frequently demanding and time-consuming, necessitating workers to pin- point problems and remember intricate procedures for their resolution in contemporary industrial environments. To address these challenges we propose an innovative assistance system that leverages a semantically zoomable Digital Twin (DT) of our factory environment to provide context-sensitive assistance to factory workers thereby helping them make informed decisions in maintenance and repair tasks. Our system aids workers by providing situation-specific guidance in an intuitive user interface in Augmented Reality (AR) glasses based on dynamic inputs such as user feedback, spatial marker registration with tracking in AR, and object detection. Based on these dynamic cues the user can zoom into the multi-layered 3D DT in Unity scene and access the necessary visualization relevant to that situation. This assistance in troubleshooting and repair procedures could potentially reduce their cognitive load and minimize time and errors. A preliminary study (N = 6) is carried out to provide an initial understanding of the impact of situation-awareness and the semantic zoom-based visualizations as assistance feature and demonstrate its usability of the assistance system.

11:00 Evaluating GPT-4o as a Cyberattack Simulator: Perspectives on AI and Human Decision-Making
Shubham Thakur, Shubham Sharma, Ranik Goyal and Megha Sharma (Indian Institute of Technology Mandi, India); Shashank Uttrani (Applied Cognitive Science Lab, Indian Institute of Technology Mandi, India); Varun Dutt (Indian Institute of Technology, Mandi, India)

Large language models (LLMs) such as GPT-4o are gaining attention for their ability to mimic human behaviors, but their use in replicating hacking strategies in the field of cybersecurity is still not fully examined. This investigation assesses GPT-4o as a simulator for cyberattacks, focusing on an important area in utilizing AI to examine the decision-making processes of adversaries. Through the HackIT simulation platform, GPT-4o's decision-making was compared against 84 human participants across bus and hybrid network topologies, under varying configurations of temperature (0.5, 1, 1.5) and top-k sampling (2, 3, 4). The results showed that GPT-4o did as well as human participants. It used an average of 32 systems in bus topologies and 28 systems in mixed topologies, and the mean squared errors (MSE) ranged from 0.03 to 1.44 in each case. In particular, the AI model was better at adapting to linear configurations and was very good at recognizing real systems, sometimes better than people. These results show that GPT-4o has a lot of promise as a defense tool for predicting and evaluating, giving us new information about how attackers work and where networks are weak. To make LLM-driven defense solutions even better, more study should be done on dynamic and adaptive attack situations.

11:20 Probing the Situational Reasoning Capabilities of ChatGPT
Andrea Salfinger and Lauro Snidaro (University of Udine, Italy)

Information exchanged in naturalistic human communication is implicitly grounded in its situational context. In particular messages exchanged via social media on on-going events, like large-scale crisis events, often assume that the actual situational context is shared by the correspondents and thus not made explicit in the message text itself. Since these messages cannot be accurately interpreted without factoring in this situational context, this renders Natural Language Processing (NLP) a challenging task in these domains. The breakthrough capabilities on fine-grained contextual understanding and Natural Language Inference (NLI) introduced with the recent Large Language Models (LLMs), however, suggest novel avenues for tackling this problem. In the present work, we thus aim to analyze current LLMs' situation understanding and situational inference capabilities, seeking to answer the question: How Well Do LLMs Understand Situational Context? We contribute i) a formalization of situational context as a conditioning factor affecting the outcome of the target task, and ii) an empirical examination of formulating this situation conditioning as a prompt engineering problem, explored on the target task of Named Entity Recognition (NER) on social media analysis for crisis computing.

Wednesday, June 4 11:50 - 13:00 (Europe/Berlin)

Lunch Break

Wednesday, June 4 13:00 - 14:30 (Europe/Berlin)

S4: Human-AI and Human-Robot Teaming

13:00 IM HERE: Interaction Model for Human Effort Based Robot Engagement
Dominykas Strazdas, Magnus Jung, Jan Marquenie and Ingo Siegert (Otto Von Guericke University Magdeburg, Germany); Ayoub Al-Hamadi (Otto von Guericke University Magdeburg, Germany)

The effectiveness of human-robot interaction often hinges on the ability to cultivate engagement-a dynamic process of cognitive involvement that supports meaningful exchanges. Many existing definitions and models of engagement are either too vague or lack the ability to generalize across different contexts. We introduce IM HERE, a novel framework that models engagement effectively in human-human, human-robot, and robot-robot interactions. By employing an effort-based description of bilateral relationships between entities, we provide an accurate breakdown of relationship patterns, simplifying them to focus placement and four key states. This framework captures mutual relationships, group behaviors, and actions conforming to social norms, translating them into specific directives for autonomous systems. By integrating both subjective perceptions and objective states, the model precisely identifies and describes miscommunication. The primary objective of this paper is to automate the analysis, modeling, and description of social behavior, and to determine how autonomous systems can behave in accordance with social norms for full social integration while simultaneously pursuing their own social goals.

13:20 Human Assessment of AI Errors and Its Impact on Hybrid Teaming for Decision Making
Ranjani Narayanan (Georgia Institute of Technology, USA); Karen Feigh (Georgia Tech, USA)

With increasing opportunities for close collaboration between humans and AI, a crucial aspect of effective teaming is humans' ability to develop mental models of their automated teammates. Particularly, the ability to discern when to accept the agents' recommendations by identifying the error boundaries and effectively overcoming brittleness. Participants in our study were tasked as disaster relief planners, assisted by a decision aid that recommends resource allocations based on several information attributes. They were asked to identify the source of the agent's errors, for low and high levels of error complexities, across which we measured team performance, participants' ability to identify and compensate for errors, and their subjective experience of working with the AI teammate. Results revealed that high error complexity degraded performance and participants' ability to form accurate mental models. While less complex errors caused satisficing among participants, high-complexity errors led to active compensation. Evaluation of their subjective experiences of working with the AI indicated that most participants underrated the agent when faced with higher complexity errors while overrating an imperfect agent for less complex errors. These findings motivate future research toward understanding how mental model formation in humans is impacted due to various levels of automation error and how that may impact human-agent teaming and collaborative decision-making.

13:40 Conceptualizing the Integration of Cognitive and Physical Models to Enable Continuous Human-Robot Interaction
Chenxu Hao (Delft University of Technology, The Netherlands); Nele Russwinkel (Universität zu Lübeck, Germany); Daniel Haeufle (University of Tuebingen, Germany); Philipp Beckerle (Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany)

Research in human-robot interaction (HRI) often puts emphasis on either the cognitive level or on the physical level. In a scenario, where a robot physically guides a person (e.g., a patient) to perform a complex series of tasks like making tea, information is exchanged on the cognitive level and forces/torques are exchanged on the physical level, continuously. Such a continuous co-adaptive interaction between both systems and the environment requires the robot to be anticipating, proactive, and able to react flexibly to the user's intentions and situation context. The unification of sequential cognitive situation modeling and continuous robotic movement control is a challenge currently missing a conceptual framework. In this perspective article, we conceptualize strategies on how to connect models of physical HRI and models of cognitive HRI, depending on the level of assistance provided by the robot system, from mere warnings of dangerous situations (level 1) to on-body continuous movement guidance (level 4). We consider the requirements for the robot to be aware of the interaction environment and have a dynamic representation of the individual user. Our perspective is intended to spark discussions and formalize assistance approaches with the aim to integrate cognitive and physical human-robot interaction approaches for anticipatory assistance in continuous dynamic tasks.

14:00 Performing Best: Organizing Suitable Human-AI Teaming for Improved Performance - an Experimental Study on Human Recognition Processes Under Time Pressure
Olena Shyshova, Audrey Kamdoum Mayouche and Mithilesh Seesunkur (University of Duisburg-Essen, Germany); Dirk Söffker (University Duisburg-Essen, Germany)

The increasing integration of advanced driver assistance systems (ADAS) and artificial intelligence (AI) in modern vehicles highlights the importance of reliable object recognition in road traffic, especially when quick decisions need to be made in difficult visibility conditions. This study investigates the human error probabilities in identifying and distinguishing objects such as cyclists, pedestrians, and traffic light colors under suboptimal conditions. It also evaluates the performance of human-AI teams in the same scenarios, comparing humanfirst and AI-first decision workflows. The results provide insights into the reliability of these approaches and offer guidance for optimizing workflows to improve road safety and efficiency for this specific example but also a clear view to the optimization of human-AI teaming.

Wednesday, June 4 14:30 - 14:45 (Europe/Berlin)

Coffee Break

Wednesday, June 4 14:45 - 16:15 (Europe/Berlin)

S5: Poster Presentations

14:45 Edge Computing Based Human-Robot Cognitive Fusion: a Medical Case Study in the Autism Spectrum Disorder Therapy
Qin Yang (Bradley University, USA)

In recent years, edge computing has served as a paradigm that enables many future technologies like AI, Robotics, IoT, and high-speed wireless sensor networks (like 5G) by connecting cloud computing facilities and services to the end users. Especially in medical and healthcare applications, it provides remote patient monitoring and increases voluminous multimedia. From the robotics angle, robot-assisted therapy (RAT) is an active-assistive robotic technology in rehabilitation robotics, attracting researchers to study and benefit people with disability like autism spectrum disorder (ASD) children. However, the main challenge of RAT is that the model capable of detecting the affective states of ASD people exists and can recall individual preferences. Moreover, involving expert diagnosis and recommendations to guide robots in updating the therapy approach to adapt to different statuses and scenarios is a crucial part of the ASD therapy process. This paper proposes the architecture of edge cognitive computing by combining human experts and assisted robots collaborating in the same framework to achieve a seamless remote diagnosis, round-the-clock symptom monitoring, emergency warning, therapy alteration, and advanced assistance.

15:00 Next-Generation Emergency Operations: Advancing Crisis Response with Drones, Augmented Reality, and Autonomous Vehicles
Swarnamouli Majumdar (Zenext AI, Canada); Biswadip Basu Mallik (Institute of Engineering & Management, India)

As global emergencies grow more complex, emerging technologies such as drones, augmented reality (AR), and autonomous vehicles (AVs) are redefining the landscape of crisis management. This paper explores the cutting-edge integration of these technologies, showcasing their potential to revolutionize firefighting, disaster response, and urban safety. Using vivid real-world examples-from drones mapping wildfires in California to AR wearables enhancing situational awareness in low-visibility conditions-this survey uncovers key applications, challenges, and societal impacts. Insights into human-centered design, trust-building, and policy evolution highlight a path forward for deploying these game-changing tools. By addressing barriers like cognitive overload, public skepticism, and ethical considerations, this research envisions a future where responders are faster, safer, and smarter in the face of disaster.

15:15 Safe and Explainable AI in 6G Telecommunication Systems: the Way Forward
Amadeu Nascimento, Junior (Ericsson Research, Brazil); Alexandros Palaios (Ericsson Research, Germany); Klaus Raizer (Ericsson Research, Brazil); Rafia Inam (Ericsson AB, Sweden & KTH Royal Institute of Technology, Sweden); Swarup Kumar Mohalik (Ericsson Research, India)

Artificial Intelligence (AI) is crucial for realizing the evolution of 6G cellular-based systems. However, the trustworthiness of AI-based models and systems - including their transparency, robustness, human agency, and safety - is equally important for rapid adoption and regulatory compliance. This paper focuses on the aspects of AI safety and explainability, and on its importance to situation awareness in future networks. We summarize the existing work, identify research gaps, and provide recommendations for integrating these aspects in a future telecommunication system, such as a 6G network. We observe that the current state-of-the-art primarily targets model-level safety and explainability, but argue that there are bigger challenges at the system-level and runtime. As we move towards 6G networks, there is an urgent need for new research directions to address these challenges. We aim to raise awareness among the research community about these critical issues.

15:30 Using Memory Contents of a Cognitive Model for Prompt Augmentation of a Large Language Model
Thomas Sievers (University of Lübeck, Germany); Nele Russwinkel (Universität zu Lübeck, Germany)

Large Language Models (LLMs) are becoming increasingly widespread thanks to their broad application possibilities and good performance. However, reliable use, for example in the provision of information, is hampered by the fact that the utterances of an LLM are occasionally inappropriate, untrue or fictitious. In addition, LLMs are limited in their ability to make human-like judgments and conclusions, especially over several steps in complex tasks. We propose an approach for augmenting the prompt used in an LLM by means of the human-like judgment and decision-making capabilities inherent in cognitive architectures for a desired deployment scenario. In particular, we access the memory contents of an ACT-R model in order to use the knowledge stored there to constrain the system prompt of the LLM in such a way that the language model can correctly reproduce facts that are otherwise unknown to it. We exemplify the use of such an approach in a Human-Robot Interaction (HRI) scenario with the social robot Pepper.

15:45 Innate-Values-Driven Reinforcement Learning Based Cooperative Multi-Agent Cognitive Modeling
Qin Yang (Bradley University, USA)

In multi-agent systems (MAS), the dynamic interaction among multiple decision-makers is driven by their innate values, affecting the environment's state, and can cause specific behavioral patterns to emerge. On the other hand, innate values in cognitive modeling reflect individual interests and preferences for specific tasks and drive them to develop diverse skills and plans, satisfying their various needs and achieving common goals in cooperation. Therefore, building the awareness of AI agents to balance the group utilities and system costs and meet group members' needs in their cooperation is a crucial problem for individuals learning to support their community and even integrate into human society in the long term. However, the current MAS reinforcement learning domain lacks a general intrinsic model to describe agents' dynamic motivation for decision-making and learning from an individual needs perspective in their cooperation. To address the gap, this paper proposes a general MAS innate-values reinforcement learning (IVRL) architecture from the individual preferences angle. We tested the Multi-Agent IVRL Actor-Critic Model in different StarCraft Multi-Agent Challenge (SMAC) settings, which demonstrated its potential to organize the group's behaviours to achieve better performance.

16:00 Goal and Strategy Evolution in Man-Machine Teams Through Dialogic Processes
Andreas Wendemuth (Otto-von-Guericke-University Magdeburg, Germany)

The collaboration of humans, intelligent physical agents (robots) and sometimes intelligent information agents (AIs, chatbots) in man-machine teams is already found in highly structured (e.g. industrial) work environments. Usually interactions take place in pairs. In this position article, a significant generalization and extension of this concept is discussed: (a) the agents are not mere tools or assistants, but proactively intervene as peers; (b) goals, strategies and actions are not completely predetermined, but evolve in the course of a dialogic process; (c) the man-machine teams are multiparty with multiple humans and intelligent agents engaging in pronounced-group appearance and modelling. Cognitive, dialogic systems form the technical basis of such team settings, which combine methods of multimodal information processing.

16:15 Listen! Audible Landmarks for Exploratory Information Retrieval in Virtual Reality
Ernst Stötzner, Juliane Höbel-Müller and Andreas Nürnberger (Otto-von-Guericke-University Magdeburg, Germany)

Supporting a user in exploring large information spaces, like collections of media objects, in order to support searching or learning about the content of a collection, is still a challenging issue. In this work, we present an exploratory information retrieval system that integrates dynamic auditory landmarks to enhance spatial orientation and thus making exploration more efficient. The system transforms a two-dimensional search map into an immersive three-dimensional environment with adaptive spatial audio cues influenced by user behavior. A preliminary user study suggests that auditory landmarks improve orientation and search efficiency, although the results are limited by the number of participants. An analysis of tracked behavior has shown promising results regarding the ability of auditory landmarks to guide the user.

16:30 Space and Air Traffic Management Situation Awareness with Notices
Erik Blasch (MOVEJ Analytics, USA)

Air traffic management has long been associated with situation awareness, especially supporting pilots in assessment and response to challenging situations. With the growth of artificial intelligence and recently large language models, multi-domain operations air and space domain operations, multi-vehicle control, and communications are embracing multi-modal fusion. In this paper, we focus on the data supporting aerospace analysis as Notice to Airman (NOTAM) and Notice to Space Operators (NOTSO). The notices afford a communication of the situation that can be extracted as an ontology to support human operators. Results show that supporting machine analytics can support operators when new situations arise.

16:45 CONSALE: CONstructing Situation Awareness in microLearning Environments
Luca Aliberti, Giuseppe D'Aniello, Matteo Gaeta and Vittorio Zampoli (University of Salerno, Italy)

The rapid evolution of digital technology necessitates quick and effective educational strategies. This paper presents the CONSALE (CONstructing Situation Awareness in microLearning Environments) project, which introduces an innovative approach to designing microlearning-based courses by integrating user-centered Situation Awareness design principles with Understanding by Design methodology. By aligning microlearning modules with explicit learning objectives and practical competencies needed in real-world settings, the CONSALE project addresses modern challenges of workplace and lifelong learning. Building on principles from Situation Awareness-Oriented Design and Goal-Directed Task Analysis, CONSALE ensures the development of coherent and focused educational units in order to support personalized and adaptive learning experiences for reskilling and upskilling across various industries.

17:00 A Situation-Aware Cloud-Edge Architecture for Cultural Heritage Management
Luca Aliberti, Giuseppe D'Aniello, Massimo De Santo and Rosario Gaeta (University of Salerno, Italy)

Cultural heritage represents a valuable asset for any Country and in particular for those with a tourism-oriented economy. The increase in tourist flows, climate change, and the natural deterioration of heritage assets make cultural heritage management increasingly challenging. New technologies can help address these challenges if appropriately designed to support the work of managers, maintenance personnel, and operators. In this context, Situation Awareness emerges as a cornerstone for designing new decision support systems and cyber-physical systems, such as maintenance systems, designed around the user to enhance the monitoring and intervention activities required. The considerable size of some cultural heritage sites, such as archaeological parks, demands pervasive and efficient monitoring solutions. The cloud-edge architecture enables the deployment of sensors and actuators at the network's edge while aggregating data in the cloud to provide a global common operating picture for decision-makers, enhancing their SA and thereby improving performance. In this article, we propose a new cloud-edge architecture model based on the principles of Situation Awareness. An illustrative example focusing on the Archaeological Park of Paestum, in Italy, demonstrates the feasibility of this solution.

Wednesday, June 4 16:15 - 16:30 (Europe/Berlin)

Coffee Break

Wednesday, June 4 16:30 - 18:00 (Europe/Berlin)

P2: Panel on Human-AI Teaming

Challenges, perspectives, and constraints in Human-AI teaming processes

Wednesday, June 4 19:00 - 21:00 (Europe/Berlin)

Conference Banquet

Cocktail and banquet.

Thursday, June 5

Thursday, June 5 9:00 - 10:00 (Europe/Berlin)

Keynote 3

Thursday, June 5 10:00 - 10:20 (Europe/Berlin)

Coffee Break

Thursday, June 5 10:20 - 11:50 (Europe/Berlin)

S6: Cognitive Modeling and Decision Making

10:20 Innate-Values-Driven Reinforcement Learning Based Cognitive Modeling
Qin Yang (Bradley University, USA)

Innate values describe agents' intrinsic motivations, which reflect their inherent interests and preferences for pursuing goals and drive them to develop diverse skills that satisfy their various needs. Traditional reinforcement learning (RL) is learning from interaction based on the feedback rewards of the environment. However, in real scenarios, the rewards are generated by agents' innate value systems, which differ vastly from individuals based on their needs and requirements. In other words, considering the AI agent as a self-organizing system, developing its awareness through balancing internal and external utilities based on its needs in different tasks is a crucial problem for individuals learning to support others and integrate community with safety and harmony in the long term. To address this gap, we propose a new RL model termed innate-values-driven RL (IVRL) based on combined motivations' models and expected utility theory to mimic its complex behaviors in the evolution through decision-making and learning. Then, we introduce two IVRL-based models: IV-DQN and IV-A2C. By comparing them with benchmark algorithms such as DQN, DDQN, A2C, and PPO in the Role-Playing Game (RPG) reinforcement learning test platform VIZDoom, we demonstrated that the IVRL-based models can help the agent rationally organize various needs, achieve better performance effectively.

10:40 Improved Situation Awareness with User Intention-Based Lane Change Assistant
Foghor Tanshi (University of Duisburg-Essen & Chair of Dynamics and Control, Germany); Dirk Söffker (University Duisburg-Essen, Germany)

Lane change assistance systems increase safety by providing warnings and other stability assistance to drivers to avert traffic dangers. In this contribution, lane change intention recognition was performed and applied to generate warnings for drivers to increase situation awareness and avoid imminent collision. The focus is to evaluate whether utilizing predictions of driver's lane change intentions to provide warnings in the event of imminent collision will result in decreased risk of accidents. It integrates an online Fuzzy-Random Forest (fuzzy-RF) approach with which collision warnings were generated. A total of 44 drivers (39 males and 5 females) among whom 22 experienced the lane change assistance system with intention recognition and corresponding warnings. The experimental and control groups were compared to determine if the intention-based warnings improve driving performance. The results indicate reduced risk of collision and enhanced performance during lane changes. In addition, the augmented reality head-up display imagery using familiar color schemes and a minimalist layout enable drivers to identify the direction or location of the potential danger without taking their eyes of the road.

11:00 Investigating the Cognitive Edge: Could tDCS Influence Curiosity and Decision-Making Under Cognitive Load?
Ankit Singh (Indian Institute of Technology IIT Mandi, India); Varun Dutt (Indian Institute of Technology, Mandi, India)

Decision-making in unknown circumstances requires balancing exploitation-leveraging past outcomes-and exploration-seeking new information-a process often constrained by cognitive load. Cognitive load may impede curiosity-driven inquiry and diminish working memory capacity, leading to less adaptive but safer selections. While previous research has shown that transcranial direct current stimulation (tDCS) applied to the dorsolateral prefrontal cortex (DLPFC) enhances cognitive flexibility and risk-taking behaviour, there is limited understanding of its efficacy in mitigating the adverse impacts of higher cognitive load on the exploration-exploitation trade-off. By examining how tDCS influences decision-making under various cognitive demands, our work fills this research gap. Under low and high cognitive load settings, 20 participants, 10 receiving tDCS (tDCS group) and 10 not receiving tDCS (control group), choose between three alternatives in a binary-choice task: safe option (Option A), risky option (Option B), and exploratory option (Option R). Results revealed that those who received tDCS were more likely to engage in exploratory behavior, choosing Option R in 40% of trials as opposed to 22% in the control group, and selected riskier alternatives (Option B) in 57.7% of trials, which was considerably more frequent than the control group (31%). Interestingly, even when subjected to a high cognitive load, tDCS individuals retained the choice for the option R, underscoring the function of neuromodulation in reducing the impact of higher cognitive load. With implications for risky situations, these findings advance our knowledge of how tDCS fosters curiosity-driven behavior through enhancement in exploratory behaviour and adaptive decision-making.

11:20 Studying Dual-Task Awareness in Industrial Settings Through Reaction Time and Physiological Signals
Abdulrahman K. Eesee (Northern Technical University, Iraq); David Kostolani (TU Wien, Austria); Vera Varga (University of Pannonia, Hungary); Taeho Kang (TU Wien, Korea (South)); Sebastian Schlund (TU Wien, Austria); Tamás Ruppert (University of Pannonia, Hungary)

Industrial environments often impose high requirements on dual-task awareness, such as simultaneously monitoring machinery signals while performing assembly work. Multitasking in such settings can strain cognitive abilities and affect work performance. This study explores attentional demands and physiological arousal in manufacturing-like dual-task scenarios. Twelve participants performed a screwing task while simultaneously reacting to a Go/No-Go test, simulating situational awareness demands in the Industry 5.0 laboratory at the University of Pannonia. The data collected included reaction times to the Go/No-Go stimuli, physiological data from the electrocardiogram and electrodermal activity (EDA), and task-related metrics such as acceleration. Our event-based analysis revealed a significant relationship between stimulus timing and the tonic component of EDA with reaction times. These findings contribute to better understanding of human factors and attention management in manufacturing settings.

Thursday, June 5 11:50 - 13:00 (Europe/Berlin)

Lunch Break

Thursday, June 5 13:00 - 14:00 (Europe/Berlin)

P3: Panel on Semantic Forensics

Semantic Forensics - Charting a path through a generative tomorrow

Thursday, June 5 14:00 - 14:30 (Europe/Berlin)

Conference Closing & CogSIMA 2026 Outlook

News

Registration is now OPEN.
Please register here!

Apr 18: Tentative program announced.
Mar 24: We are please to announce that Prof. Nicole Krämer (University of Duisburg-Essen, DE) will present a keynote address.
Mar 10: Prof. Hasan Ayaz (Drexel Univ., USA) will give a tutorial on Neuroergonomics for Situation Awareness.
Mar 7: Mike Kozak (Lockheed Martin Advanced Technology Laboratories, USA) will give a tutorial on Detecting Generated and Manipulated Media Using Semantic Forensics.
Mar 6: Prof. Galina Rogova (Univ. of Buffalo, USA) will instruct a tutorial on Decision Making under Risk.
Mar 4: Prof. Antonio Lieto (Univ. of Salerno, IT) will present a tutorial on Cognitive Systems Design
Dec 13: CogSIMA 2025 group rates for accommodation announced.
Dec 11: In response to numerous requests, we grant a final paper submission extension to Dec. 20.
Dec 9: Registration fees have been posted.
Nov 27: We are delighted to share that Prof. Niels Taatgen (Univ. of Groningen, NL) will present a keynote address on The Human Engineer's Toolbox.
Nov 8: We are excited to announce that Prof. Niels Taatgen (Univ. of Groningen, NL) and Prof. Nicole Krämer (Univ. of Duisburg-Essen, DE) will present keynote addresses.
Aug 6: We are deeply saddened by the passing of our CogSIMA 2024 General
Co-Chair Dr. Melita Hadzagic
.

 

Sponsors and Patrons

IEEE logo
 
SMC logo

University of Duisburg-Essen logo

ISIF Logo

Related Conferences & Organizations