Aside from recent, spirited public debates over the role of AI in defense, the broader and near-ubiquitous role of software in defense has received less attention. Credit: public domain.
Unlike the commercial sector, the defense industry’s most dominant contractors occupy the same established position they have held for the last three decades. While it remains to be seen whether this will hold, software’s work “under the hood” has transformed their world in three key ways.
In 2002, as they prosecuted a war against Al Qaeda, senior leaders in the U.S. national security firmament followed the motto “The world is a battlefield.”1 Almost a decade later, Silicon Valley venture capitalist Marc Andreessen penned a now widely cited Wall Street Journal op-ed describing disruption from U.S. software companies in other industries, aptly titled “Why Software Is Eating The World.”2 If the world is a battlefield and software is eating the world, the logical syllogism would hold that software is also eating the battlefield.
But while software’s role in modern military operations has increased drastically, it has not yet transformed the defense industry as it has retail, mobility, or hospitality. Tom Goodwin’s 2015 TechCrunch article gave an insightful observation of software’s effect in the commercial space. “Uber, the world’s largest taxi company, owns no vehicles,” he wrote. “Facebook, the world’s most popular media owner, creates no content. Alibaba, the most valuable retailer, has no inventory. And Airbnb, the world’s largest accommodation provider, owns no real estate.”3 But in defense, the most dominant players are still, at least for now, the established prime contractors that have dominated for the last three decades. Just how will software eat the defense industry itself is the question of the day. The answer is not yet clear, but three broad categories stand out: production of weapons, military operations themselves, and distribution of those systems to the theaters.
Software’s Long March Through the (Defense) Institutions
This is not to say the defense industry has been left untouched by the waves of technological transformations of the past two decades. Remotely-piloted aircraft, for instance, have become a central feature of armed forces, providing persistent surveillance over land and sea, compressing the targeting-strike cycles from weeks to seconds, and reshaping close-range combat, as the experience in Ukraine shows. Along the same lines, SpaceX’s re-usable and more efficient launchers, together with smaller and cheaper satellites, have permitted a dramatic growth in the number of space launches as well as of satellites in space. Starlink as a result is able to provide widespread connectivity, also for military operations. Similarly, digital platforms such as Discord, Telegram, or X represent some of the primary sources for open-source intelligence world-wide.
Yet with the exception of vague and often highly-politicized discussions about artificial intelligence, the role of software in defense has received relatively less attention from the broader public, despite the fact that it has come to play an increasingly critical role.4 In abstract terms, software can affect any industry at three different levels: how its products are made, how they are used, and how they are moved to those arenas. It is worth a brief exploration of each to assess how software can be expected to develop next.
Software’s Role in Production
First and perhaps paramount, software supports and streamlines the production of weapon systems. Software ensures superior operational performance by replacing human beings for certain tasks, thus yielding faster and more accurate results. Computer-assisted design, computer-assisted manufacturing, computational fluid dynamics, finite element analysis, and increasingly also additive manufacturing, play an essential role.
It is worth noting in this regard that the defense industry has played a key role in the emergence of computer-assisted manufacturing, owing significantly to the stringent technical requirements the U.S. Air Force and Navy demanded of its prime contractors in the days after World War II. Without computer numerical control (CNC) machine tools, the high-precision machining would not have been possible.5 From assistance in manufacturing, software then started to take an increasing role also in other stages of the production process, primarily design, thus relieving weapon designers from the need to carry out complex calculations about multiple trade-offs.
Software then expanded to assessing the technical properties of the alternative shapes and materials used, thus reducing testing during the design and development stages.6 The emergence of 3D printing is nothing else than a further evolution of computer-assisted design and manufacturing (CADAM), for components that meet a specific parameters (e.g., complex geometries, low production runs, etc.).7 Finally, software has come to play an increasing role also in maintenance, maximizing cycles between repairs by helping to identify the most efficient time for replacement and repair thus reducing ground-time.
Software’s Role in Operations
Possibly even more dramatic is the role that software has come to play in military operations, from tasks such as flight control of highly maneuverable aircraft; detection, identification, tracking, and engagement of hostile targets (the “kill-chain”); data-fusion and ocean surveillance; as well as electronic warfare and counter-measures.8 With improvement in aerodynamics and propulsion made possible by software, the performance of aircraft reached a point where human beings could no longer, by themselves, pilot an aircraft because of the need to for micro-seconds adjustments based on external data gathered and processed by multiple sensors.9 Further improvements in aircraft called for additional software support, including for “mundane” aspects such as regulating oxygen provision, as well as for more specific ones, such as the combat system aggregating together multiple information about the external environment.10
This point leads to the central role that software has come to play in addressing one of the longstanding problems facing armed forces more generally: surveillance of large geographical areas, especially oceans.11
With the development of steam engines and turbines at the end of the 19th century, which allowed for superior speed, ocean surveillance came to pose serious problems for navies. This only became more demanding with time, as the operational need for an accurate and updated assessment of enemy positions became more difficult and demanding. By relying on multiple sources of information, including human intelligence, electronic intelligence (capturing radio emission), intercepting and deciphering radio communications, and more, this problem could, in theory, be addressed. But the need to aggregate multiple sources of time-sensitive and at times contradictory information to create a unified “picture” of the battlefield was a nightmare that military and civilian personnel could no longer address unaided. The number of sensors and information had increased, and the number of potential incongruences among different sources (e.g., the same ship being labeled with two or more different names) had also increased. Through data-fusion, software has come to represent a real solution, as it could organize all the incoming information into a complete, time-sensitive and accurate picture – even though the process has been slow and marked by problems and setbacks.12
Along the same lines, software has come to play a central role in detecting, identifying, tracking, and (eventually) engaging enemy targets.13 Media and policy attention often focuses on some attributes of modern missiles, such as the long-range and high-precision, as the debate about the HIMARS and ATACM missiles provided to Ukraine shows. Yet, often missing from these discussions is the fact that, without the capacity of accurately detecting, identifying and geolocating/tracking enemy targets, long-range high-precision missiles are essentially useless. The same applies to air defense, which has received significant attention because of the remarkable performance in the war in Ukraine, with the alleged interception of multiple Russian aero-ballistic missiles and in the Middle East, with the interception of Iranian ballistic and cruise missiles. It goes without saying, software (precisely, signal processing and machine learning) is behind all these tasks and successes.
Software’s Role in Distribution
In a world of finite resources, with constraints of time and space making trade-offs all the more consequential, software also plays a key role in getting the right tools to the right place and at just the right time. Detection, identification, and tracking certainly play a critical role as hostile forces try to conceal their presence in order to approach their target undetected.14 But software also is crucial in moving the necessary systems to deter and counter those forces. It computes a large number of statistical calculations that help, first and foremost, in monitoring a specific area (whether the ocean surface, ocean depth, airspace, or ground) to detect anomalies and facilitate efficacious distribution of the systems needed to engage them.15 Software can accurately identify a given object by looking for a potential match of its profile (whether visual, acoustic, electromagnetic, or otherwise) in a library of stored profiles (i.e., exactly as the app Shazam does).16 Finally, software can help track and engage an enemy by observing the direction and speed of a given object and predicting its immediately subsequent position so that it can provide the necessary information to a missile, which will then use the physical features of the object (whether emitted, such as heat, or reflected, such as radar or laser reflection) for terminal homing.17
Conclusions
Over the past 25 years, the commercial industry has gone through a massive transformation, which is evident when we look at the largest companies by market capitalization. In 2000, the behemoths of the second industrial revolution like Exxon, General Electric, and Shell were at the peak of American capitalism (together with companies like Microsoft and Cisco). In 2024, the largest companies by market capitalization are all in the high-tech: Apple, Amazon, Meta, and Nvidia.
In contrast to the dynamics of the commercial industry, defense, with a few notable exceptions, is still dominated by prime contractors like Boeing, Lockheed Martin, and Northrop Grumman. Such stark differences, however, should not lead to the unwarranted conclusion that software has not had a tectonic impact in the defense industry. It has. A potential explanation for the difference in trends between the commercial and the defense industry might stem from the fact that many defense companies pioneered themselves the development of software. The question, however, is whether they will be able to keep pace with the advances of their children.
The experience of the automotive sector, in this regard, is telling as some established car makers, most notably Volkswagen, struggled with software development. Admittedly, Volkswagen did not have an established experience with software, which is not the case for defense companies. But for them, the twin problem will be in recruiting and retaining top talent in an increasingly competitive market, and in being able to upgrade its software experience at a time of exponential growth.
Along the same lines, whereas a century ago, companies like Ford, FIAT, Bristol Aeroplane Company, ThyssenKrupp, and Renault transitioned in a few years from commercial producers to defense manufacturers, we have not seen a trend of software-centric companies acquiring the size of top prime contractors or companies like Amazon or Google becoming key weapons producers – yet.
More than 20 years after the September 11th attacks, the world is, once again, a battlefield. It is one in which combatants fight bit by bit as well as bullet by bullet.
References
1 Jeremy Schall, Dirty Wars: The World Is A Battlefield (New York: Nation Books, 2013).
2 Marc Andreessen, “Why Software Is Eating The World,” Wall Street Journal, August 20, 2011.
3 Tom Goodwin, “The Battle Is for The Customer Interface,” TechCrunch, March 3, 2015.
4 For a notable, albeit niche exception, see for example, Christine Anderson and Merlin Dorfman, eds., Aerospace Software Engineering: A Collection of Concepts (Washington, D.C.: American Institute of Aeronautics and Astronautics, 1991).
5 J. Francis Reintjes, Numerical Control: Making a New Technology (Oxford, U.K.: Oxford University Press, 1991); Roberto Mazzoleni, “Innovation in the Machine Tool Industry,” in Sources of Industrial Leadership: Studies of Seven Industries, ed. David C. Mowery and Richard R. Nelson (New York: Cambridge University Press, 1999).
6 David E. Weisberg, “The Engineering Design Revolution: CAD History,” shapr3D, December 12, 2022.
7 Ibid.
8 Frank Heilenday, Principles of Air Defense and Air Vehicle Penetration (Washington, D.C.: George Washington University, 1988); Electronic Warfare Fundamentals (Nellis Air Force Base, Nev.: Air Combat Command Training Support Squadron, 2000).
9 James E. Tomayko, Computers Take Flight: A History of NASA’s Pioneering Digital Fly-By-Wire Project (Washington, D.C.: National Aeronautics and Space Administration, 2000).
10 Steve A. Fino, Tiger Check: Automating the U.S. Air Force Fighter Pilot in Air-to-Air Combat, 1950–1980 (Baltimore, Md.: Johns Hopkins University Press, 2017); Timothy P. Schultz, The Problem with Pilots: How Physicians, Engineers, and Airpower Enthusiasts Redefined Flight (Baltimore, Md.: Johns Hopkins University Press, 2018).
11 See, for example, Norman Friedman, Network-Centric Warfare: How Navies Learned to Fight Smarter Through Three World Wars (Annapolis, Md.: Naval Institute Press, 2009).
12 Ibid.
13 Simon Kingsley and Shaun Quegan, Understanding Radar Systems (Edison, N.J.: SciTech, 1999); J. C. Toomay and Paul J. Hannen, Radar Principles for the Non-specialist, 3rd ed. (Edison, N.J.: SciTech, 2004).
14 Filippo Neri, Introduction to Electronic Defense Systems (London: Artech, 2018).
15 Mark Denny, Blip, Ping, and Buzz: Making Sense of Radar and Sonar (Baltimore, Md.: Johns Hopkins University Press, 2008).
16 Oleg I. Sukharevsky, ed., Electromagnetic Wave Scattering by Aerial and Ground Radar Objects (Boca Raton, Fla.: CRC, 2014).
17 William G. Ballard and Stéphane Kemkemian, “Fire-Control Radar,” in Principles of Modern Radar: Vol. 3: Radar Applications, ed. William L. Melvin and James A. Scheer (Edison, N.J.: SciTech, 2014).