Future Land Warfare Collection 2021: Data to Decisions: Enabling Military Decision-Making through Operations Research and Systems Analysis

The commander must work in a medium which his eyes cannot see, which his best deductive powers cannot always fathom; and with which, because of constant changes, he can rarely become familiar.[1] 

Carl von Clausewitz (1780–1831)

In early June 2020, the Australian Strategic Policy Institute published an article titled ‘Australia’s Navy needs operations research to navigate the future’.[2] The authors’ central argument posits that operations research (OR) is the critical capability required to enable the Royal Australian Navy (RAN) to adopt a data-driven analytical approach to decision-making.[3] Their position is that OR is essential for identifying, achieving and sustaining critical capabilities as economically as possible in a post-COVID-19 resource-constrained environment.[4] This situation is neither unique to the RAN nor confined to the area of capability development and management. It is a common situation that spans the entire Australian Defence Force (ADF) system of systems—strategic, operational, tactical, capability development, preparedness, and people.

Data-driven, analysis-based decision-making is the standard today in most successful businesses. Chief data officers and analytic staff with data scientists or operations analysts are increasingly a part of their structure.[5]

The Australian Government’s Force Structure Plan 2020 outlined a total package of capability investment in land forces of approximately $55 billion over the next decade, representing 20 per cent of government’s total capability investment in Defence over this period.[6] This will enable land forces to meet challenges into the future, with new investments in long-range strike weapons, watercraft, helicopters, information effects, logistics resilience, and emerging robotics and autonomous systems. Traditionally, professional military judgement has been a primary method to answer questions regarding the acquisition, associated force design, and operational employment of new capability. However, military commanders are increasingly seeking scientific answers to these problems that fuse quantitative analysis and qualitative insights to improve operational effectiveness and capability decision outcomes.[7]

The future operating environment is likely to be characterised by previously unimagined levels of convergence, data saturation and speed. Military commanders must be able to efficiently and effectively leverage large volumes of quantitative data to make informed, timely decisions to achieve successful mission outcomes. The military that can best capture, process, and analyse data effectively is likely to have a marked advantage over strategic competitors.[8] OR cells across the entire ADF, not just within the RAN, that provide quantitative (modelling and data analysis) and qualitative (operationally relevant insights) decision support will be the critical capability that enables this. This essay explores the origins and operational significance of military OR, discusses the application of OR in a broader ADF context, and addresses some general considerations for the growth of an increased ADF OR workforce. 

What Is Operations Research?

OR within Defence is defined as ‘The analytical study of military problems undertaken to provide responsible commanders and staff agencies with a scientific basis for decision on action to improve military operations’[9]—the application of the scientific method to decision-making[10] or, more succinctly, data to decisions. The formal discipline of OR grew from the efforts of military planners, mathematicians, and statisticians working in multidisciplinary teams during World War II.[11] In the years since, OR has grown into an active field of academic research, with methods, models and tools permeating from the military enterprise into business, government, and society more broadly.[12]

Employing a range of techniques from across many fields, OR is inherently multidisciplinary and leverages mathematical and computer sciences such as probability-based modelling, statistical analysis, optimisation, large-scale simulation, machine learning and data science. It fuses these fields to optimise solutions, helping organisations and decision-makers understand themselves better and operate better.

OR practitioners are commonly known as operations research analysts, operations analysts, industrial engineers or, within the ADF, operations research systems analysts (ORSAs). These professional analysts are master problem-solvers. They decompose problems into their essential elements, determining the right permutation of methods, models and tools to identify and analyse the causal issues of complex applied problems.[13]

The key value proposition for Army’s continued investment in a uniformed ORSA workforce is the organisational control and enhanced flexibility options it offers. Beyond support to force design and capability development, ORSAs offer a unique ability to support operational decision-making. The ORSA workforce is a persistent uniformed analytic capability capable of providing real-time quantitative decision support to operational commanders. Army’s ORSAs provide analysis not only for seeing what exists and what has happened (descriptive analytics) but also for reliably predicting what is likely to happen (predictive analytics), as well as the intelligent generation of reliable courses of action (prescriptive analytics). Examples of the value of this to Defence include the design and analysis of the 2020 Force Structure Plan experimentation, COVID-19 modelling and forecasts, Army Operational Force Structure supply and demand analysis, and operational optimisation studies.

History illuminates many cases where the use of OR could have avoided disaster. A case in point is Operation Eagle Claw, where the use of OR may have led to a vastly different outcome. In November 1979, Iranian revolutionaries breached the American Embassy in Tehran, capturing 52 embassy staff. Five months into their captivity, the United States attempted a rescue operation, codenamed Operation Eagle Claw. The after-action review[14] uncovered several planning challenges with the mission design. It found that a major factor contributing to mission failure was the insufficient number of helicopters used.

Military planners had determined that they would need a minimum of six helicopters to evacuate the hostages. The planners used a simple form of average-based decision-making to determine the number of helicopters required. They did not apply probabilistic assessment to understand the likelihood of platform failure, particularly that arising from helicopter reliability. As a minimum of six helicopters was required to complete the mission, planners identified the need for eight helicopters to start the mission. They determined this by using an average platform availability percentage, approximately 75 per cent for this particular helicopter. 

Figure 1: The aftermath on the ground of the failed hostage rescue[15]

Average-based decision methods do not take into account the range of possible outcomes or their associated likelihoods, offering only a single point to represent something far more nuanced. In this instance, an OR analyst may have suggested a technique such as Monte Carlo simulation, or provided insight on possible mission outcomes using a model such as binomial distribution. Based on the insights from either of these methods, an OR analyst would have advised that there was only a 68 per cent likelihood of completing the mission with at least six helicopters when starting with eight. To ensure a 95 per cent mission success likelihood based on helicopters alone, the OR analyst would have advised starting the mission with at least 11 helicopters, as described in Figure 2.

Figure 2: Probability of successfully having six or more helicopters to complete the mission[16]

This example highlights what having an ORSA in the planning team could provide, helping to ensure that decision-makers have access to better methods than average-based calculations. ORSAs use analytical methods and mathematically based procedures to enable leadership decisions in a continually changing environment. They introduce rigorous quantitative and qualitative analysis to support military decision-making processes. They work in diverse disciplines that include personnel management, doctrine and force development, training management, system testing, system acquisition, decision analysis, and resource management, as well as tactical, operational and strategic planning from brigade level through to the highest levels of the Department of Defence.

How Could Operations Research Be Applied to ADF Operations in the Future?

The ADF’s uniformed OR capability is less formalised than that of some of our partners, such as the US Department of Defense (US DoD), where professional communities exist as a part of specialist career stream models. Presently the ADF has nine qualified analysts across the enterprise; only two have contemporary experience and are employed in that role.[17] This difference manifests for the US DoD in the way in which senior leaders are supported, with ORSAs embedded in cells across their highest-level commands. The ADF presently has a single cell in an equivalent structure with uniformed analysts: the Land Warfare Laboratory in Future Land Warfare, supporting the Head of Land Capability in Army. Access to similar types of analysis by other senior leaders within Defence comes from a mix of Defence Science and Technology Group, academia and industry, all of which offer a range of different employment profiles to uniformed analysts.  

An ever-increasing number of Defence’s senior leaders have had exposure to the US system. The integration of Australian officers in coalition headquarters has consistently demonstrated the value of embedded OR analyst cells, not only for enterprise functions but also to deliver time-sensitive, resource-constrained analysis in support of operations. Uniformed ORSAs provide essential decision support to senior leadership decision-making. Their deep technical expertise and operationally relevant experience are the basis for this. 

Defence has a range of positions across the enterprise codified[18] broadly for OR, operations analysis, and operational assessments. However, sufficiently qualified or experienced personnel are rarely employed in these roles. Moreover, a critical mass of professional uniformed ORSAs does not exist within Defence. The requisite skills in advanced OR, applied mathematics and computer science are not invested in by Defence to the degree necessary to address these known workforce gaps or increase the Defence OR workforce. 

The Defence Science and Technology Group maintains an OR capability within the Joint and Operations Analysis Division. However, it is resource constrained to meet the full spectrum of roles and tasks that the ADF requires. To fill this identified gap, Defence requires uniformed ORSAs to complement and collaborate with those in the Defence Science and Technology Group, focusing on a nuanced spectrum of capability, preparedness, operations and strategic analysis.  

The following vignette demonstrates the value of an OR cell to an operational organisation such as Joint Operations Command. In early 2017, a three-person OR cell comprising a United States Navy officer, a United States Marine Corps officer and an Australian Army officer completed a short-term study for the Commander of the United States Pacific Fleet.[19] The team investigated how strategy and tactics could assist de-escalation in the South China Sea while maintaining coalition strategic influence and ensuring that essential trade flows remain uninterrupted.

This team conducted the study with five key considerations: tactical, logistical, geographical, diplomatic and political feasibility. They designed an analysis framework using an advanced network optimisation model to confirm the strategic constraints and identify tactical employment locations of land-based weapon systems. The cell modelled critical chokepoints to determine the optimal placement, constrained by the number of weapon systems available to forces in the region. The analysis informed understanding and enhanced the decision-making process for principal military commanders. It directly informed concept of operations development, as well as theatre strategy for military integration with other government agencies.

Figure 3: A visualisation of shipping data which informed the quantitative analysis

Concept development and operational problem-solving through quantitative analysis are not only possible but also realistic now to support ADF joint operations. Applied OR such as this acts as a catalyst for conversations to evolve how joint operations are planned and executed to meet the challenges of the future together.

How Could the ADF Generate the Right Operations Research Capability?

The employment of ORSAs across the ADF necessitates a teaming approach. ORSAs must be paired with an experienced military planner and augmented with subject matter experts as required and determined by the field of analysis. The main advantage of this approach is the blending of the quantitative (ORSA) and qualitative (military planner), addressing complex problems from multiple perspectives. This pairing forms the core group that must be augmented with domain expertise relevant to the field of inquiry. 

Presently the ADF educates its uniformed OR analysts only at the Naval Postgraduate School (NPS), a first-class graduate research institution and home to one of the world’s leading OR programs.[20] Australia has no advanced education program delivered within a military context that could produce an equivalent ORSA as the NPS program does. Australia does have world-class universities with programs that can replicate the range of skills required, albeit across several independent graduate-level programs. Currently there are no single multidisciplinary programs, even outside a military context, for the ADF to qualify ORSAs in Australia. 

OR analyst workforce development[21] is a problem that is systemic throughout our coalition partners; however, the ADF does face two unique challenges. First, the ADF does not currently have sufficient throughput of suitable candidates to qualify as OR analysts. Second, the ADF does not currently have the institutional academic organisations, with residential experience, capable of replicating a full military OR program similar to those of our coalition partners. Notwithstanding this, the ADF could generate sufficient OR analyst effect through a mixture of multidisciplinary quantitative skill sets, created in partnership with our domestic academic institutions. 

Coalition senior executive the late Kevin E Williams, formerly Director U.S. Air Force Studies, Analyses and Assessments (AF/A9), defines three key areas for address increasing complexity and uncertainty in strategy and operations:[22] inform leaders; strengthen analytic communities to meet the challenges of tomorrow; and improve (decision) clarity. He asserts that we are at an inflection point for OR. Accordingly, the US Air Force has designed a five-part strategy to address this, consisting of intentional, tailored education, training, and professional experience; accessible, referenced data; collaborative, interdisciplinary methods; simple, modular, and scalable models and tools; and a dynamic, collaborative community of analysts and partners. Williams describes the output of this approach as the hyper-enabled analyst. The purpose of this new approach is to optimise the scarce analyst resource to better support senior leader decision-making. 

The ADF could adopt a model similar to that of the US Air Force to grow the OR workforce, particularly the education model. Evolving the current approach of qualifying all ORSAs at NPS, the ADF could generate complementary quantitative analysts across a range of STEM fields, such as computer science, applied mathematics, statistics, and engineering. This change would require dedicated organisational investment in both cultural change and foundational education, recognising the value of technical qualifications to the future ADF workforce, as identified in the Defence STEM Vision Strategy.[23]

OR is an essential element of the STEM Vision Strategy and must be prominently featured. The strategy document lays out the vital elements to promote and retain an advanced STEM workforce, explicitly noting the requirement for STEM tailored career pipelines, talent management, sustainable workforce growth, and understanding the key attributes required for a STEM career. The foundational requirement to realise this capability is a centrally managed career model, which is systemic across groups and Services in Defence.

Conclusion

Dedicated and planned investment in the future of Army’s ORSAs is important as, given the ever-increasing demand for analytic skills across Defence, the ADF will necessarily need to ‘raise, train and sustain Operations Research / Operations Analysis skills and gain practical experience applying analytical approaches to military problems’.[24] The military that can best capture, process and analyse data effectively is likely to have a marked advantage over strategic competitors. Uniformed ORSA cells across the entire ADF to provide quantitative (data analysis) and qualitative (operationally relevant insights) decision support will be the critical capability that enables this. 

A critical short-term priority for OR development should be support to operations. Uniformed ORSA cells organic to Joint Operations Command and all joint task force headquarters should be considered the minimum requirement to support principal commanders’ decision-making adequately. In the longer term, all functional commanders across the ADF should have access to uniformed ORSA cells to support decision-making. 

Any investment to increase the ADF’s uniformed ORSA capability will significantly improve the quality of support to senior leader decision-making. While we have outlined some broad considerations for ORSA workforce development, these represent just a few of many potential options. Further analysis and planning are required to develop a mature understanding of the demand and approach needed to fulfil this. 

The authors wish to thank Jeffrey E Kline, Mary Hill and Arthur H (Trip) Barber for their review, commentary and insights for this essay. This paper was first written in August 2020.