Real-Time Human-In-The-Loop Simulation with Mobile Agents, Chat Bots, and Crowd Sensing for Smart Cities.

Area of Science:

• Socio-technical systems
• Computational social science
• Agent-based modeling

Background:

• Social interaction and network modeling are crucial across disciplines, from social sciences to smart city management.
• Current virtual simulations often lack real-world diversity, relying on artificial behavior or offline data.
• Smart cities require adaptive control using real-world sensor data, with humans acting as potential sensors.

Purpose of the Study:

• To introduce a framework integrating agent-based simulation with crowd sensing and social data mining.
• To bridge the gap between virtual simulations and real-world complexity by augmenting virtual environments with live data.
• To enable real-time interaction between simulated and real-world elements.

Main Methods:

• Development of a framework combining agent-based simulation with crowd sensing using mobile agents.
• Utilizing mobile agents for crowd sensing through chat dialogues (chat bots) and mining smartphone sensors.
• Creating augmented reality/virtuality by linking simulated and real-world interactions.

Main Results:

• Demonstrated the suitability of augmented agent-based simulation for social network analysis.
• Showcased the use of parameterized behavioral models and mobile agent-based crowd sensing.
• Validated the framework through three distinct use-cases.

Conclusions:

• Augmented agent-based simulation offers a powerful methodology for studying complex socio-technical systems.
• Mobile agents and crowd sensing effectively bridge virtual and real worlds, enhancing data diversity.
• The framework provides a robust approach for real-time analysis and control of large-scale social dynamics.