Getting good results is only half the journey
Defence scientist Katie Mortimer-Murphy's data analytics toolbox helps tell the whole story.
While contributing to several critical Defence workforce analysis studies, DSTG aerospace engineer and mathematician Katie Mortimer-Murphy has been researching and developing data and visual analytics enhancements for DSTG’s ATHENA workforce simulation tools.
While her double degree in aerospace engineering and mathematics touched on basic aspects of data visualisation, Katie has developed an interest in this area since joining DSTG’s workforce analysis team. ‘It’s clear that a big part of ATHENA’s success is the visual way that the output is communicated to people. I can see how important it is not just to get good results but to explain those results.’
Over several years, Katie has been developing ways to use data analytics to get a more useful understanding of what’s going on in a workforce, and to better present results to Defence clients. Her prototype toolbox is currently being integrated with the ATHENA Lite tool, and will be available for use by Defence workforce analysts and planners later this year.
Katie shared her achievements at Australia’s Defence Operations Research Symposium (DORS) last December. Her talk “Simulation and data analytics for a Defence workforce transition” was selected as the winner of the Gus Schaefer Award for best paper from DSTG.
‘My DORS presentation was about my toolbox, which builds on the existing ATHENA simulation tool created by our team, and how I used it in relation to help Army with a Battlefield Aviation workforce analysis task. I also covered the “design of experiments” aspect, running simulations using the toolbox to generate a whole bunch of data, and then the various data analytics and interactive visualisations I created to communicate the results. I built a user interface that allows clients to interact with the mathematical “heat maps”.’
‘Our team’s aim is to make these complex analytical concepts approachable for people who don’t have a numerical modelling background; most people will understand Bayesian network analysis when it’s visualised.’
Mathematics + military know-how = more, together
While our defence scientists are skilled at applying specialist knowledge to perform data analytics, it’s important to use insights from military subject matter experts to guide experiments. ‘Otherwise, you’re coming up with solutions that might not be feasible or useful to Defence,’ says Katie. ‘By displaying our science in a better way, we are able to combine the best of science and military wisdom to achieve more.’
‘ATHENA Lite is already deployed on Defence networks and used by Defence planners. My user interface and toolbox takes the huge amount of simulation data that is generated when analysing workforce transitions, and visualises it.
‘For example, we’ve been helping Army plan a transition from the ARH helicopter to the Apache helicopter*. It’s a big transition that needs careful planning. You can’t just take the helicopter capability off-line for six months for retraining. When we first started this workforce planning there was not even a clearly defined new platform. That’s when visualising lots of simulations with lots of variable inputs can really help planners.’
The Gus Schaefer Award includes funding for Katie to travel to the 90th MORS (Military Operations Research Society) Symposium later this year in the United States, at the Marine Corps University in Quantico, Virginia, to present her work in a special session. If international travel opens up enough for her to take the trip, she also plans to visit the Canadian defence science team at DRDC which has been sponsoring ATHENA for several years. Katie has monthly progress meetings with the Canadians, who are starting to use ATHENA for some of their own Canadian Defence workforce planning. The trip to Ottawa would be a valuable opportunity to showcase DSTG’s ATHENA progress and see firsthand how an international defence organisation is using it.
*Defence project LAND4503 Phase 1 ARH replacement.