In May 1953, in the desert west of Idaho Falls, a crew powered up the world’s first naval nuclear propulsion system. What made it possible was daringly aggressive innovation: Adm. Hyman Rickover insisted that the Submarine Thermal Reactor Mark I be built to exactly the specifications that would later be required inside a submarine. This meant hundreds of pounds of simulated sea pressure per square inch, shock resistance tested to the standards of a depth-charge attack, and air conditioning sized at three times the requirement. Rickover’s own engineers fought him, arguing that the basic challenge of building a nuclear reactor at all was hard enough without simultaneously solving submarine problems too. After all, this was one of the first-ever experiments on practical applications of atomic power. Rickover overruled them to demand a realistic operational prototype.
When Mark I reached full power in June 1953, Rickover ordered his engineers to test the system on a simulated transatlantic crossing at full power. The test was unplanned. It was also opposed by his senior officer on site and the technical officers at the Naval Reactors Branch, who called from Washington and urged him to stop. But when Rickover refused and completed the test, which took over 65 hours, he had successfully shown the ability for a nuclear-powered submarine to cross the Atlantic nonstop at full power without surfacing. Rickover faced considerable resistance from Navy leadership, and the pursuit of nuclear propulsion made many career diesel submariners uncomfortable. The proposal had initially sounded, in the words of Rickover’s project officer, “like a trip to the moon” to most who heard it. But when the test was complete, skeptics had nothing left to argue.
Sixty years later, the same logic played out on a launchpad. For decades, the Air Force treated reusable rockets as an engineering fantasy. The aggressive counterargument came in December 2015, when SpaceX’s Falcon 9 first stage returned to Cape Canaveral and landed itself vertically, engines burning, on four deployable legs, eighteen stories tall and intact. The reusable booster landing was the culmination of a bet that the aerospace establishment had repeatedly pronounced irrational. Today, SpaceX holds eighty percent of the global launch market. That fantasy rocket has eaten the industry.
During the first wave of air strikes in Operation Epic Fury on Feb. 28, 2026, one of the most effective tools cost $35,000 per unit and had been in the U.S. arsenal for roughly eight months. The Low-Cost Uncrewed Combat Attack System was built by SpektreWorks, an Arizona startup that began by recreating the same airframe as an Iranian Shahed-136. They were awarded a contract, not through a program office, but by the Accelerate the Procurement and Fielding of Innovative Technologies program administered by the Office of the Under Secretary of Defense for Research and Engineering. Their $30 million contract and seven-month timeline from unveiling to combat would not have been possible through a traditional program of record.
The history of defense innovation is full of cases like this: where a leap-ahead capability arrived from a direction that nobody had considered and was nearly disregarded because it didn’t look like the existing program of record. For the reactor in the desert, the reusable rocket booster, or the drone design lifted from an adversary, the capability had to be prototyped and proven to operational readiness before institutions would accept it. In each of these cases, the project was not pre-aligned to existing programs or procurement budgets. Program offices tend to be unprepared for innovation like this, due to calcified assumptions about what is feasible, after spending decades and billions of dollars on mature legacy systems. And when this kind of innovation comes along, inflexible budgets limit what is investable.
A New Funded Pathway for Innovation Insertion
For decades, the U.S. “Triple Helix” innovation ecosystem has engaged industry, academia, and government together to discover and fund new technology in concert. Small Business Innovation Research and Small Business Technology Transfer programs fund small companies, universities, and labs through competitive awards. The Defense Advanced Research Projects Agency seeds speculative science at the technical frontier. The Defense Innovation Unit scouts and contracts commercial solutions. Services like the Space Force operate front door offices to engage industry. But none of this innovation infrastructure addresses the transition: the moment when a validated technology tries to move from a prototype into operational use. That gap, known since the 1990s as the “valley of death,” persists in part because innovation funding is appropriated for exploration while program budgets, which should go to operations and scaled production, are locked years in advance. The space between them has no natural funding home. Rickover had Atomic Energy Commission funding. SpaceX had Elon Musk’s personal capital and found a commercial market willing to pay for launches before the Air Force was. SpektreWorks had the Accelerate the Procurement and Fielding of Innovative Technologies research and engineering program. When a prototype demonstration is done and the system is proven, the hard question becomes how and where to scale, operate, and sustain the capability.
A new acquisition mechanism might be the answer. Beginning in fiscal year 2028, each portfolio acquisition executive (the successor to program executive offices, responsible for managing and fielding entire capability portfolios rather than individual programs) will be required to include an innovation insertion increment in their budget. As discussed in a previous article on this site, portfolio acquisition executives currently lack the flexible capital to absorb new commercial solutions even when those solutions have been demonstrated. The innovation insertion increment is designed to fix that. Whether it does will depend on implementation choices being made right now.
Done right, the innovation insertion increment is transition capital: money held specifically for inserting validated innovative technology at speed. It is not a research fund, and it is not a requirements generator. Its purpose is to give portfolio acquisition executives a mechanism to say yes when an innovator presents a new capability and asks for co-investment or follow-on support. The technology should already have been matured outside the portfolio (whether through Small Business Innovation Research, Defense Innovation Unit, the Defense Advanced Research Projects Agency, or private capital) and should arrive with an operational prototype rather than a pitch. Innovators bear a reciprocal obligation: to do the derisking work in advance, so that a portfolio acquisition executive can issue a defensible investment decision rather than an exploratory research bet. The innovation insertion increment provides the capital for what comes after demonstration is complete.
In practice, this means innovation insertion increment funding should deliberately bias nontraditional vendors without an established contracting relationship with the portfolio, bringing technology that is genuinely new to the portfolio and has already absorbed private or other government innovation investment. These criteria are necessary conditions that keep the innovation insertion increment aligned to its purpose. The Department of Defense’s program-centric structure systematically
The Department of Defense’s program-centric structure systematically defaults to incremental improvements on legacy systems rather than genuine technology insertion. Without guardrails in place and enforced, portfolio acquisition executives are likely to direct their new flexible capital toward existing programs of record, merely creating a slush fund and defeating the mechanism’s intended purpose. In the face of cost overruns and entrenched priorities, “rapid capability insertion” can easily get redefined as accelerating the delivery of increment six of a system that entered development in 2009 with no innovation involved. The history of discretionary innovation accounts, including the Rapid Innovation Fund and various reprogramming and transfer authorities, is one of flexible dollars getting absorbed into legacy program baselines, with transition metrics that were either absent, poorly tracked, or defined in ways that failed to capture whether genuinely new capabilities reached the field.
The same memo that established portfolio acquisition executives contains a possible accountability mechanism: Each portfolio acquisition executive should publish a portfolio scorecard and have it tracked through monthly acquisition acceleration reviews chaired by the under secretary of defense for acquisition and sustainment. The primary metric for these scorecards is currently delivery speed, but other metrics could be included to reveal whether innovation insertion increment dollars are reaching new capabilities versus simply lubricating existing programs. For example, scorecards could track what share of each portfolio acquisition executive’s increment went to vendors without an existing contracting relationship with the portfolio, what share went to nontraditional contractors, what share leveraged private capital investments, Defense Innovation Unit, or small business funds. Congress could go one step further by requiring this innovation insertion increment-specific reporting as a condition of the Fiscal Year 2028 appropriation. That kind of accountability would make the use of this capital visible and attributable. Evaluating and reporting these metrics can help shape decisions and orient portfolio acquisition executives’ incentives.
Recognizing innovation requires portfolio acquisition executives to define their portfolios around capabilities rather than the systems they already own. The Mark I reactor, the Falcon 9, and the Low-cost Uncrewed Combat Attack System drone did not look like their predecessors, but were new modes to deliver undersea power, space launch, and air strike. GPS is not the only way to deliver position, navigation, and timing to a warfighter in a GPS-denied environment. Therefore, a portfolio acquisition executive responsible for positioning and navigation has a mandate that extends beyond the system currently on contract. They should be asking what puts a soldier or a vessel on the map when GPS is unavailable, and stay open to new answers that do not look like what they already manage. This is the kind of disposition that the innovation insertion increment is designed to enable, and one that institutional defaults work against.
The innovation insertion increment, implemented with fidelity to its purpose, creates a funded pathway for the next act of daringly aggressive innovation to move from operational prototype to sustained capability. The people writing the implementation guidance today have an opportunity to build a genuine bridge across the valley of death.
Isobel Porteous is head of business development at EarthTraq Corporation. She studied aeronautics and astronautics at Stanford University, where her thesis on space arms control verification won the William J. Perry Prize. The views in this article are her own and do not represent those of EarthTraq or any other organization.
**Please note, as a matter of house style, War on the Rocks will not use a different name for the U.S. Department of Defense until and unless the name is changed by statute by the U.S. Congress.
Image: Cpl. Kayla Mc Guire via Wikimedia Commons
