Image source: B-21 Raider, Secretary of the Air Force Public Affairs
On December 1, 2022, US Secretary of Defense Lloyd Austin published a memo establishing the Office of Strategic Capital (OSC), meant to provide necessary capital to help aspiring defense companies bridge the so called “valley of death,” the significant time gap between the development of a successful prototype and awarding of a government production contract. This time gap has contributed to the Pentagon’s struggle to incorporate emerging technologies, leaving it to prod along with acquisitions that are often relegated to minor purchases or surface-level improvements on existing platforms, such as swapping out the targeting system in the M1 Abrams Main Battle Tank—the model which first came into service in 1985 and remains in service today.
On December 2, the Department of Defense (DoD) revealed the B-21 Raider, the United States’ first new strategic bomber in three decades. In the same span of time, technology companies such as Google, Apple, Microsoft, and many others have changed our everyday lives, commercializing the global positioning system, personal computers, smart phones, and advanced search algorithms. These emerging technology firms adhere to a new business methodology known as ‘lean startup’ and are evolving their products at a pace and scale that dwarfs the traditional defense sector. Unless the DoD embraces the principles of lean startup to reform its antiquated acquisition processes, the United States will be at risk of losing its military superiority which has been an important pillar of the world order since the end of the Cold War.
In 2016, Krause Schwab, executive chairman of the World Economic Forum, announced the arrival of the fourth industrial revolution, “characterized by a fusion of technologies that is blurring the lines between the physical, digital, and biological spheres.” Technologies such as artificial intelligence (AI), wireless connectivity, robotics, and three-dimensional printing are all evolving exponentially rather than linearly, while simultaneously converging upon one another. The fourth industrial revolution has the potential to revolutionize military warfare and provide increasingly lethal capabilities to increasingly smaller entities. The technological convergence of robotic and wireless technology has enabled drones, and improvements in AI will enable fully autonomous, networked operations of these systems. Three-dimensional printing will allow the manufacturing of repair parts or entire replacement drones right on the battlefield. Equally important, these technologies are becoming cheaper and more accessible, leading to their rapid diffusion. The first drone strike occurred in 2001, in a US operation targeting Taliban operatives in Yemen. In the two decades since, at least 80 countries now have unarmed drones and 24 have armed drones. The combined speed of technological progress and diffusion of capabilities will require the U.S. to embrace modernized innovation and acquisition practices now in order to adapt to a more rapidly lethal changing future.
Commercial Innovation
The commercial sector is leading the charge on technological innovation in the twenty-first century, from AI, robotics, and three-dimensional printing—all dual-use technologies not limited to the military. Innovations of previous decades were mainly incubated, developed, and deployed by governments and their militaries, but, in the 1980s, venture capital started to replace government funding for investments in companies that were spearheading emerging technologies. The focus of these companies subsequently shifted from seeking government contracts to pursuing profit. Commercial technology is now progressing at a rate that governments and modern militaries simply cannot maintain pace with. The F-35 Lightning II, the world’s most advanced fighter jet, costs over $100 million per unit and has a computer system that can process 400 billion operations per second; meanwhile, Nvidia, a commercial technology company, can produce the Nvidia DRIVE AGX Pegasus, a graphics processing unit that can conduct 320 trillion operations per second and used to develop autonomous driving for cars or trucks. In other words, Nvidia has a chip that is 800 times faster than what is present on the F-35.
US adversaries like the People’s Liberation Army (PLA) have recognized this shift toward commercial innovation and have adapted new strategies to leverage those strengths. In 2017, Xi Jinping established the Central Military-Civil Fusion Development Commission, which focuses on incorporating commercial technologies into the PLA. Civil-Military Fusion enables the Chinese Communist Party to leverage all Chinese commercial firms and their associated innovations, intellectual properties, algorithms, and data collected from consumers to the benefit of the central government. Although the PLA’s innovation strategy has been regularly dismissed to be mainly absorptive, leaning on espionage, copyright theft, and reverse-engineering, the viability of Civil-Military Fusion should not be dismissed in light of the rapid innovation now occurring within the commercial sector. The advent of rising global Chinese technology companies like Alibaba, Tencent, Huawei, and ByteDance (owner of TikTok), can provide the leg-up needed for the PLA to leapfrog its technological development.
DoD Innovation
Currently, the DoD acquisition process is perfectly suited for procuring complex platforms that require a decade of design, development, and financing—all requiring stable conglomerates and entrenched legislation. During the 1990s, the US defense industry underwent drastic consolidation and shrank the numbers of prime contractors from 51 to five, resulting in the current crop of the ‘primes’ or the five largest US defense contractors: Lockheed Martin, Boeing, General Dynamics, Raytheon, and Northrop Grumman. Primes have the overwhelming advantage in experience, relationships, understanding of how to influence Congress, and familiarity with retaining DoD as a loyal customer. In exchange, Congress is reluctant to make changes to the Federal Acquisition Regulation (FAR) and the Planning, Programming, Budgeting, and Execution (PPBE) process, which regulate DoD acquisitions and is criticized as being confusing, complicated, and cumbersome—often requiring expensive consulting services to navigate the process.
This convoluted system of ingrained relationships, processes, and legislation renders it virtually impossible for new entrants, like Silicon Valley technology companies, to sell their products and services to the government. The DoD recognizes the importance of improving its relationship with Silicon Valley and has made recent attempts to connect Palo Alto to Washington DC. Within the last decade, the Pentagon established new organizations like the Defense Innovation Unit (DIU) and the Defense Innovation Board to conduct outreach and to streamline the adoption of innovation throughout the national security enterprise. This has inspired the service branches to follow suit, and there are now over 35 different organizations to onboard disruptive technologies into the federal government. These organizations have also been successful at leveraging existing legislative authorities such as Other Transaction Authorities (OTAs) as a workaround to the stringent regulations of the FAR by funding prototype development and getting new companies interested in initiating business relationships with the DoD.
However, despite the DIU’s inception in 2015, successful prototypes almost never translate to long-term production contracts with the DoD. The 2021 Annual DIU report states that 279 cumulative prototype contracts have been awarded while only 35 of them have successfully transitioned to a production contract. The valley of death is a phenomenon so termed due to aspiring technology companies having to wait 12-24 months of open competition despite successfully providing a prototyped solution to a government-identified problem. In the fast-paced world of Silicon Valley startups, this time period is too long for a promising start-up to remain stagnant when it could be raising capital and refining its product against real customer feedback. To their credit, organizations like the DIU have become proficient at identifying companies with promising technological solutions and then providing the initial funding via OTAs to rapidly prototype. However, once the government proves unable to provide additional funding to sustain the initial success, companies become incentivized to take their prototypes to the commercial sector and to seek additional funding through venture firms or foreign investors. Additionally, as technology evolves, the requirements of two-years ago can quickly become obsolete, competitors can find ways to interject themselves to disrupt the process, and the DoD is forced to adopt an outdated capability or cancel the bid before restarting the process. Most notably, the Pentagon cancelled its 2018 JEDI contract, a $10 billion Pentagon award to move its information technology system to the cloud, due a feud between Amazon and Oracle. Pending the outcome of the Joint Warfighting Cloud Capability (JWCC) contract, the Pentagon’s data is still not available on the cloud, and the DoD is unable to take advantage of cloud technology’s enormous cost savings, improvement in data security, and better resilience in data storage—despite nearly every Fortune 500 company doing so.
Embracing Lean Startup
Silicon Valley startups follow the so-called lean startup method, favoring experimentation, customer feedback, and iterative design. The method embraces the opportunity to fail rather than shunning it. The traditional business model of writing a five-year business plan, pitching it to investors, assembling a team, introducing a product, and finally selling it currently results in a 75 percent failure rate. In many ways, the current DoD acquisition process perfectly mirrors this traditional process. In comparison, the lean approach has proven to be a far quicker and cheaper method to launch products that customers actually want.
The lean startup method has three key principles: 1) write a framework known as a business model canvas; 2) listen to customers through a process called customer development;and 3) practice agile development. A business plan requires months of planning and research and results in an untested hypothesis, while a business model canvas enables flexibility and provides the intent on how a company plans to create value. In customer development, a startup undergoes business experimentation to seek a business model that works. This prevents overinvestment in a process or a product that has no product-market fit and allows the company to quickly test different hypotheses. Finally, agile development enables the product to be developed iteratively and incrementally by constantly incorporating customer feedback throughout the process. This reduces the time and resources necessary to find a minimum viable product.
The US national security enterprise and the OSC should tap into the knowledge and experience left behind by the failures of thousands of startups. First, especially in the beginning phases, the OSC should eliminate strictly defined requirements in the Request for Proposal and replace them with a general hypothesis or problem statement. This will allow for product flexibility and open the aperture to a broader array of innovative solutions. Furthermore, this will provide wiggle room for if and when senior DoD leaders revise the request, which can be frustrating for small companies as it stands. Second, prototypes need to be produced at a scale that can be quickly tested by the end-user, most likely the warfighters. This will enable companies to make the best improvements from data provided though real-world use. One way to implement this is by fielding prototype capabilities to US Army brigades that are preparing for a rotation through a Combat Training Center. If fielded to a unit early enough in the preparation phase, a productive cooperation resembling customer development could form between the warfighter and the prospective vendor. Third, the acquisition process should adopt agile development to enable companies to iterate on the product alongside warfighter feedback. This will require the DoD to have longer-term, multi-year, variable-sized production contracts with a wide variety of potential suppliers and, in essence, choosing some of the potential winners.