Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC) 2015
2015 Paper No. 15165 Page 1 of 11
Modelling a helicopter training continuum to support system transformation
Michael Johnstone, Vu Le, Burhan Khan,
Doug Creighton
Ana Novak, Vivian Nguyen,
Luke Tracey
Centre for Intelligent Systems Research
Defence Science and Technology Organisation
michael.johnstone@deakin.edu.au,
vu.le@deakin.edu.au, burhan.khan@deakin.edu.au,
dougc@deakin.edu.au.
ana.novak@dsto.defence.gov.au,
vivian.nguyen@dsto.defence.gov.au,
luke.tracey@dsto.defence.gov.au.
ABSTRACT
This study investigates the role of system dynamics (SD) modeling to support strategic decision making for an aviation training
continuum that is going through major change. The Australian helicopter training continuum (HTC) is currently undergoing
transformation, with restructure and consolidation of training schools and training platforms across multiple services. In this
research, we introduce a novel SD-based HTC simulation architecture to facilitate the discovery of relationships between
student and instructor development and flow dynamics. The proposed simulation architecture employs hybrid push – pull flow
control to quantify transience and estimate recovery time after a policy change or disturbance. This architecture allows for
multiple student and instructor types, and their respective intake levels and pass rates. Here the instructor variables include
availability, specialization and experience. Enos (2011) successfully explored the application of SD modeling to understand
the behavior for combat aviation training in an individual school. This research employs a similar modeling philosophy, but
takes a higher level view of the system by looking across multiple training schools, which introduces complexity due to pooling,
latency and the amplification of affects across the system. The ability to identify causal relationships allowed stakeholders to
develop a deeper understanding of the underlying systemic problems, such as delayed transitions between schools and instructor
shortages, whilst the hybrid “push-pull” design allowed us to quantify the pooling of students between schools.
ABOUT THE AUTHORS
Michael Johnstone is a Senior Research Fellow within the Centre for Intelligent Systems Research. He holds a Ph.D. degree
in simulation-based learning from Deakin University in 2010. His research interests lie in simulation modeling, scheduling and
multi-objective optimization for intelligent decision support systems.
Vu Le is a researcher at the Centre for Intelligent Systems Research. He graduated as a Mechanical Engineer (Honours) at
RMIT University. He hold a M.E. degree in the area of multi-line multi-product capacitated lot-size with workforce production
planning and also obtained a Ph.D. degree in modeling, simulation and analysis of complex conveyor networks from Deakin
University. His research interests include optimization, modeling simulation and data analysis of industrial systems.
Burhan Khan is an associate research fellow at the Centre for Intelligent Systems Research. He holds a Masters of Engineering
from Deakin University (2010). His research interest lie in many-objective optimization, scheduling, machine learning and
software design.
Doug Creighton is Associate Professor in Systems Engineering and Deputy Director of the Centre for Intelligent Systems
Research. He obtained a Ph.D. degree in simulation-based optimization from Deakin University, Australia, in 2004. His
research interests include modeling, simulation, optimization, visualization, and decision support.
Ana Novak is an Operations Research Scientist for Defence Science and Technology Organization (DSTO) in Australia. She
received her B.Eng. (Honours) in Information Technology and Telecommunications from University of Adelaide in 2003. She
obtained a Ph.D. degree in Mathematics from The University of Melbourne, Australia in 2006. Her research interests include
queuing theory, probabilistic graphical models, human and computer vision, Bayesian inference modeling, simulation,
optimization, and decision support.
