580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can multi-scale and multi-agent simulation frameworks be integrated to effectively model complex, heterogeneous dynamic systems?
This research area investigates the formulation of simulation frameworks that can address the complexity of dynamic systems exhibiting multiple spatial and temporal scales, heterogeneous agents, and integrated domain-specific behaviors. It matters because many real-world systems—from urban mobility to ecological and economic systems—require modeling interactions across diverse scales and agent types to generate accurate, actionable insights. Effective integration of multi-scale simulation with agent-based approaches enables simulations that capture emergent phenomena arising from interactions at different levels, supporting better decision-making and policy analysis.
Key finding: SimMobility integrates short-term (micro), mid-term (meso), and long-term (macro) simulation levels into a unified platform under the activity-based modeling paradigm, enabling agent decisions spanning milliseconds to years... Read more
Key finding: Introduces PADAWAN, a formal agent-based framework for multi-scale simulation that uniformly represents nested environments and agent multi-membership in spatiotemporal scales through an interaction-oriented approach (IODA).... Read more
Key finding: Proposes a grid-based computational infrastructure to support large-scale multi-agent simulations involving millions of autonomous agents distributed over dynamic environments with limited communication capabilities. It... Read more
Key finding: Presents MARS, a cloud-based simulation platform designed to enable Modeling and Simulation as a Service (MSaaS), supporting large-scale agent-based models with seamless integration of GIS and time-series data. MARS provides... Read more
Key finding: Develops a conceptual framework advocating natural solver-based approaches (e.g., cellular automata, genetic algorithms, lattice Boltzmann methods) for efficient mapping of complex systems to massively parallel computing... Read more
2. How can agent-based simulation frameworks be designed to effectively represent social behaviors and interactions in discrete event environments?
This theme focuses on the development of discrete event social simulation (DESS) frameworks and agent behavior models that faithfully capture the timing, sequencing, and interactions of social processes. This line of research is crucial for understanding complex social and behavioral dynamics relevant to public policy, military decision-making, and human factors engineering. Incorporating discrete event simulation concepts into multi-agent systems facilitates rigorous temporal management of social events, enhances modeling of actor states and communications, and supports the analysis of potential alternative social futures.
Key finding: Proposes a framework embedding multi-agent systems within the discrete event simulation paradigm to represent complex social behaviors and interactions. The approach leverages discrete event simulation concepts—such as event... Read more
Key finding: Provides foundational principles for conducting simulation studies, emphasizing the dual nature of simulation as both an art and science, particularly highlighting the importance of carefully designing input data, models, and... Read more
Key finding: Introduces Agent.GUI, a multi-agent simulation framework built on JADE that enhances usability and accessibility for experts unfamiliar with agent technology, through a customizable graphical user interface. It extends JADE... Read more
3. What advances enable the composability, interoperability, and reuse of simulation models via ontologies and standardized modeling components?
This research theme explores the development of semantic frameworks and XML-based standards to support the composability, interoperability, and reuse of simulation components across heterogeneous simulation engines and domains. Ontology-based representation of simulation models and standards such as Base Object Models (BOMs) enable semantic integration, facilitate querying, comparison and inference on simulation artifacts, and promote modular construction of simulations. These advances matter for improving the efficiency and effectiveness of simulation development and execution in multidisciplinary contexts.
Key finding: Proposes an ontology-based approach for representing domain-specific simulation models as instances of simulation engine ontologies, enabling semantic interoperability, querying, inference, and comparison across heterogeneous... Read more
Key finding: Describes the Base Object Model (BOM) concept, an XML standard that captures reusable simulation pattern components—Interface BOMs codifying interaction patterns and Encapsulated BOMs modeling entity behaviors—to enable rapid... Read more