AsiaSim 2013: 13th International Conference on Systems Simulation, Singapore, November 6-8, 2013. Proceedings

Gary Tan, Gee Kin Yeo, Stephen John Turner, Yong Meng TeoNational 2013 · DOI

Introduction

The proceedings of AsiaSim 2013, the 13th International Conference on Systems Simulation, document a significant epoch in the evolution of modeling and simulation (M&S). Held in Singapore from November 6–8, 2013, the conference curated 63 peer-reviewed papers—comprising 45 full and 18 short contributions—that collectively define the frontier of the field during that period. The volume reflects a strategic pivot toward managing extreme system complexity through advanced parallelization and interdisciplinary applications.

The Challenge of Systemic Complexity

The core challenge addressed by the AsiaSim 2013 contributors is the widening "complexity gap." As real-world systems—from biological pathways to global socio-technical networks—become increasingly interconnected, traditional linear modeling techniques often fail to provide the necessary fidelity or speed. The proceedings identify three critical hurdles:

  1. Scalability: The requirement to leverage multicore and distributed computing clusters for massive-scale simulations.
  2. Fidelity vs. Performance: The persistent trade-off between high-accuracy modeling and the need for real-time execution.
  3. Interdisciplinary Expansion: The migration of simulation methodologies into non-industrial domains such as public health (e.g., obesity epidemics) and social behavior (e.g., crowd dynamics).

Methodological Innovations

Two primary methodological pillars emerged as dominant themes within the conference:

Three-Level Parallelization

A key architectural framework proposed by keynote speaker Yiping Yao involves a hierarchical approach to parallelization designed to maximize hardware utilization:

  • Job-level (Multi-sample): Executing independent simulation samples in parallel to explore parameter spaces.
  • Task-level (Multi-entity): Utilizing Message Passing Interface (MPI) to manage complex interactions between discrete simulated entities.
  • Calculation-level: Accelerating intensive mathematical models within individual entities using GPGPUs, FPGAs, and Many Integrated Core (MIC) architectures.

Agent-Based and Hybrid Modeling

Agent-Based Simulation (ABS) remains a cornerstone for modeling emergent phenomena from the bottom up. Furthermore, the proceedings emphasize hybrid and co-simulation techniques, which allow for the integration of disparate software environments (such as ADAMS and MATLAB) through High-Level Architecture (HLA) gateways, enabling more comprehensive system-of-systems modeling.

Key Insights: Natural Information Processing

A transformative perspective offered during the conference suggests viewing complex systems as "natural information processors." By treating phenomena like viral spread or social networking as computational processes, researchers can bridge the gap between theoretical deductivism and data-driven inductivism. This paradigm shift encourages the identification of "signatures of order" within seemingly chaotic data.

Results and Impact

The conference maintained high academic standards, with a ~47% acceptance rate for full papers (45 accepted out of 95 submissions). The research presented demonstrated significant progress in "latency-hiding" schemes for distributed clusters, which address communication bottlenecks in parallel environments. Additionally, the inclusion of "serious games" for professional training—such as the SanTrain system for emergency medical response—indicates the growing maturity of interactive simulation platforms.

Conclusion and Future Directions

AsiaSim 2013 signaled the internationalization of the Asian simulation community, moving for the first time outside the "Golden Triangle" of China, Japan, and Korea to Singapore. Looking forward, the proceedings point toward "Socio-dynamics" and "Multi-scale applications" as the next frontiers. These areas will require even more robust integration of formal methods with computational modeling to validate "soft" social models against real-world observational data, mirroring the rigor currently found in physical engineering simulations.

Summary updated

BibTeX
@inproceedings{tan2013asiasim,
  author = {Gary Tan and Gee Kin Yeo and Stephen John Turner and Yong Meng Teo},
  title = {{AsiaSim 2013: 13th International Conference on Systems Simulation, Singapore, November 6-8, 2013. Proceedings}},
  booktitle = {National},
  year = {2013},
  doi = {10.1007/978-3-642-45037-2},
}
Figure