Recent history has provided mixed messages on military modernization and whether nations should invest in next-generation technology. For instance, the United States and its NATO partners provided the Afghanistan National Army (ANA) with material support years beyond what the insurgents had. Yet, the ANA quickly collapsed without coalition support. Conversely, Russia suffered greatly after failing to modernize its military to counter the threat of newer anti-tank weapons and unmanned aerial vehicles. Given the differences between the conflicts, it may seem unfair to compare the two, except that in each conflict, success or failure depended upon how people made the most of the equipment available. Many leaders immediately turn to material solutions to outpace an adversary’s modernization, thinking each situation is like Russia’s failure to modernize. However, when it comes to effectively using data to make better decisions, the Army is more akin to the ANA, and fielding more technology may not be the right application of resources.
Soldiers in the U.S. Army bring an intuitive understanding of technology that transfers to the employment and sustainment of modern weapon systems. This familiarity stems from dealing with advanced machinery and technology on a daily basis, something many ANA soldiers lacked. Because of this unfamiliarity, many ANA vehicles and weapons went without preventive maintenance. Training and additional systems to track maintenance helped, but the ANA soldiers often reverted to their initial behavior and understanding. It became evident a culture of preventive maintenance required changes in their education and development.
The Army is attempting to modernize all things data after seeing how civilian organizations have benefited from incorporating data analytics into their processes. However, the average Soldier may not intuitively understand data analysis and the mathematics that support predictive logistics. This has led many within the military to grasp onto hype surrounding automated data analytics, especially artificial intelligence (AI) and machine learning, without fully understanding the technology, as if the object provides an advantage over an adversary. This may be hard for many to accept, but the similarities between handing the ANA a helicopter and providing Soldiers with automated data analysis are eerily similar. Developing a culture that embraces data-driven decision-making and quantitative reasoning takes years of education.
Many civilian organizations have centralized departments dedicated to data analytics, meaning most managers do not require indoctrination into a data culture. Why does the Army require decentralized data analysis and more emphasis on leaders at all levels to be educated? The short answer is the Army’s mission sets are more variable, multifaceted, and broadly defined, while most civilian companies have a single objective — maximize profit. Civilian companies have well-established, unchanging systems that benefit from data gathered from sensors. At the same time, the military often changes established relationships and processes based on mission requirements, with most data being provided through human input. The ability to conduct tailored, decentralized operations has been a hallmark of the Army’s success on the battlefield, but it makes centralized data analysis difficult. A deliberate effort must be made to educate all Soldiers on data literacy, with leaders obtaining a greater understanding of how to use data to inform decisions. A culture of quantitative reasoning will not be created quickly, but it starts with delivering the education many Soldiers are not getting in high schools and colleges.
Modernizing information technology and data analysis capabilities are vital to maintaining the lead over near-peer adversaries. However, the Army must be cautious not to allow the technology to outpace its understanding of how to employ and maintain it effectively. While advances in automated decision tools have been impressive, there is a serious risk of them being biased toward the conditions that were used to develop the algorithms that made them successful. An AI tool trained on logistic data from the National Training Center will perform poorly if blindly applied to a unit training at the Joint Readiness Training Center. Even with more mundane analysis tools, correlation can be mistaken for causation, leading to poor decision-making. Without a concerted effort to overhaul ingrained behaviors while educating servicemembers on the foundations of data analysis, leaders will resort to leaning heavily on experience and the art of decision-making while giving less consideration to the science of decision-making. Before the Army invests heavily in modernizing its data systems, it must invest in modernizing its education systems to ensure its people know how to fight and win with data.
William T. Smith, Ph.D., is an instructor of operations research for the Army Sustainment University at Fort Gregg-Adams, Virginia. He has more than 29 years of military experience as a logistics officer and operations research analyst. He earned a Master of Science in applied mathematics from the Naval Postgraduate School and a Doctor of Philosophy in industrial engineering from Pennsylvania State University.
Editor’s Note: This article was a selection from the Army Sustainment University President’s Writing Competition.
This article was published in the Summer 2023 issue of Army Sustainment.