2002/3 and 2003/4 Annual Reports
This research suggests developing novel control-theoretic
concepts and techniques for overcoming the core problems encountered in
managing supply-chains associated with discrete manufacturing. The outcome of this research expected to be a
theory and technology applicable to semiconductor manufacturing and other types
of discrete-part manufacturing problems. In particular, to apply control
approaches developed for the continuous flow processes in supply-chain
networks, it is necessary to overcome fundamental issues regarding continuous
to discrete mappings and contextualization of concepts, scale, and efficiency
of the problems encountered in the supply chain domain. The research suggests
developing a framework to simulate supply chains for both the prediction need
by the tactical decision “inner” loop and the validation of the outputs from
the “outer” strategic inventory and capacity planning systems. The research
inquires are:
§
the extension of control-oriented approaches
based on Model Predictive Control (MPC) to
a novel class of problems in discrete parts manufacturing that extend beyond
chemical process control and traditional formulations of MPC,
§
the effective use of bi-level nonlinear
programming to accomplish systematic tuning of MPC controllers in an uncertain
environment, and
§
the development of a
software architecture that can enable the unique computational needs of supply
chain class of problems through simulation modeling.
Engineering of large-scale, complex systems (e.g., semiconductor supply-chain network) requires simulation modeling. To employ a simulation and modeling framework, it is important for it to offer principles and methods for not only modeling and simulation, but also explicitly enabling validating and verifying simulation models. This research extends the DEVSJAVA modeling and simulation software architecture to support automatic validation and verification in general and semiconductor supply-chain network in particular. The research inquiries are:
§ Software Architecture Specification and Development for non-invasive simulation experiments
§ Scaleable modeling of supply-chain process models with built-in control and command features
§ Multi-Formalism Modeling Approach for Semiconductor Supply/Demand Networks
Studying long-term human activities such as landuse and
agropastoral has been the key to earth’s landscapes and society. Understanding
the interplay among humans, land, climate, and tools is expected to provide
fundamental understanding enabling investigation and therefore prediction of
human activities and landscape. This multidisciplinary research focuses on
archeological and paleoecological study of the Mediterranean spanning 8000
years beginning from the early Holocene period. With the availability of recent
advances in geospatial modeling and agent simulation in combination with
archeological field data, this study investigates the prehistoric
§ the effects on biodiversity of growth in agropastoral systems,
§ the changing impacts of landuse intensification and diversification on landscapes, their resilience, and vulnerability to degregation, and
§
the long-term
sustainability of human maintained socioecosystems in varying environmental and
social contexts.
Modeling of decentralized network systems based on ecologically-inspired communication and control patterns support handling design specification. System designs benefit from the decentralized inner-working of biological systems such as bee colonies. This research develops an environment for studying complexity and scalability of network system topologies and protocols under varying experimental settings. The research questions under investigation are:
§ Developing model specifications based on dynamic and adaptive swarm-based routing protocols
§ Extend the DEVSJAVA environment to support studying alternative routing algorithms such as shortest-path and distance vector.
ACIMS entered into a long term relationship with Northrop Grumman Information Technology to support modeling and simulation at the Joint Interoperability Test Command. ACIMS will undertake tasks as requested throughout the lifetime of the NGIT contract for Joint Distributed Engineering Plant and other projects, a long term developmental effort. ACIMS will supply research and development expertise and services to support NGIT in line with the following major themes:
§ Testing and evaluation in DoD’s systems acquisition process
§ Distributed Simulation for System Testing
§ Formalized Model Development
§ High Performance Simulation
§ Education and Professional Development
See link for more information.
This project developed the DEVS (Discrete Event System Specification) Formal Framework for Scalable Enterprise Design and extended earlier-developed DEVS-based modeling and simulation environments.
See link for more information.
§
Founding
member and active participation in the M&SNet
activities such as expansion of the M&SNet member organizations and development
of collaborative research proposals
§
Hosted and
organized the first M&SNet meeting
in
§
Hosted
and organized the NSF Workshop on Modeling and
Simulation for Design of Large Software-Intensive Systems: Challenges and New
Research Directions
§
ACIMS
and ASU launched an Online
Master of Engineering in Modeling & Simulation
§
Founding
editor of SCS Journal of Defense Modeling and
Simulation sponsored by AMSO