The Valley of Death Is Real

Why promising technologies stall between TRL 3 and 6, and what the data says about fixing it.

Here Are the Numbers.

The United States invests over $140 billion annually in defense research, development, test, and evaluation alone. Billions more flow through civilian agencies and university research programs. Yet a persistent pattern emerges across every sector: promising technologies stall between Technology Readiness Levels 3 and 6.

The gap is so well-documented it has a name. The valley of death.

Industry estimates suggest four in five new inventions are never commercialized. Academia focuses on TRLs 1 through 4. Private investors do not engage until TRL 7 or above. The space in between, TRLs 4 through 6, is where technologies go to die. Not because they lack technical merit. Because the system is not built to carry them across.

This is not a failure of imagination or talent. It is a structural feature of how the U.S. innovation system is designed, funded, and incentivized.

Research Mandates End Where Commercialization Begins Federal laboratories and universities are funded for discovery, not deployment. Their budgets are tied to generating new knowledge, not manufacturing products or validating markets.

The Department of Defense structures its RDT&E budget across distinct activity codes. Codes 6.1 through 6.3 fund the science and technology base. Codes 6.4 and above fund advanced development and prototyping that moves toward acquisition. These categories operate under separate authorities, planning cycles, and risk tolerances.

The practical result: most federally funded research reaches TRL 3 or 4, a validated proof of concept in a laboratory environment, and then hits a wall. Advancing further requires systems engineering, integration testing, pilot production, and field validation. These activities typically cost between $2 million and $10 million per technology. They fall outside the mission and budget authority of research organizations.

Universities face the same constraint. Technology transfer offices typically file patents and license IP at TRL 3 to 4, when proof-of-concept data exists and publishable results are available. Holding IP longer and maturing it further would require prototyping capital, dedicated development teams, and business infrastructure that universities are not built to provide.

The numbers bear this out. UT Austin generated 300 invention disclosures in FY2025, up 15% from the prior year. The University of Florida registered 446, a record. The University of Minnesota launched 26 startups. Disclosure volume is not the bottleneck. The bottleneck is what happens next.

Incentives Reward Discovery, Not Deployment The incentive structures within research institutions reinforce the plateau.

Academic researchers are evaluated on publications, grants, and scientific novelty. Federal laboratory scientists operate under similar frameworks. None of these reward systems prioritize manufacturing readiness, regulatory clearance, supply chain validation, or market adoption, the exact capabilities required to move a technology from TRL 4 to TRL 6 and beyond.

The Government Accountability Office has repeatedly found that the gap between development and acquisition communities within DoD impedes technology transition. The acquisition community requires higher levels of maturity than research organizations are resourced to deliver.

The median royalty rate on university-licensed technology is roughly 2% of product sales. The TTO typically receives about 25% of that. For a technology transfer office with a $10 million annual budget, the university needs $2 billion in cumulative licensed product sales just to break even. The vast majority of patents are never licensed. When they are, a single license often accounts for nearly all of a TTO's income.

Revenue is the wrong metric. But it is the one most institutions measure. The result is a system that optimizes for filing patents at TRL 3, not for building commercialization pathways at TRL 5.

Budget Architecture Creates Structural Friction In defense, Research, Development, Test, and Evaluation dollars fund early-stage work. Procurement dollars fund production. These appropriations flow through different accounts, are planned years in advance, and cannot be easily transferred.

DoD manages over 1,700 RDT&E and Procurement budget accounts. Reprogramming funds between them requires approval from OSD, OMB, and all four congressional defense committees. Program managers hesitate to sponsor technologies at TRL 5 or 6 unless a program of record exists, requirements are locked, and a funding line has been identified. Without that alignment, promising technologies stall regardless of technical merit.

The SBIR program illustrates the challenge at smaller scale. Phase I awards of roughly $150,000 to $275,000 fund feasibility studies aligned with TRL 3 to 4 work. Phase II awards of about $1 million require evidence of a commercialization strategy and transition pathway. Phase III, the stage where technologies reach operational use, has no dedicated federal funding at all. The company must independently secure a contract or customer.

The overall Phase I success rate across all agencies averages 15 to 20 percent. Even for companies that win, the timeline from submission to funding runs 6 to 12 months depending on the agency. Many companies submit multiple proposals before landing their first award. The system is designed around patience and persistence, not speed to market.

The Missing Middle The core bottleneck is not any single program or funding mechanism. It is the absence of a stakeholder who owns the transition from research to product.

Laboratories generate scientific options. Entrepreneurs generate products. But the capital, expertise, and operational focus needed to bridge that gap must come from somewhere. Customer discovery. Market validation. Transition pathway development. Capital strategy. Federal research grants are not structured to cover this middle layer at scale. Venture capital typically does not engage until TRL 7 or above, when technology risk has been substantially retired.

The valley of death describes a critical 18 to 24 month transition period. Technologies that have achieved proof of concept require tens of millions in capital and specialized operational expertise to reach commercial scale. Organizational inertia, not technical failure, determines outcomes.

Recent institutional responses reflect growing recognition of the problem. The DoD established the Office of Strategic Capital in 2023, with congressional authority to issue loans of $10 million to $150 million in critical technology areas. The Defense Innovation Board's 2023 report, "Terraforming the Valley of Death," concluded that the problem is not the level of investment but how investment dollars are deployed.

DOE's Technology Commercialization Fund invests millions annually to bridge the gap between lab research and commercial application. In FY2025, TCF funded 42 projects totaling $35 million in direct funding with $21 million in partner cost share. The FY2026 CLIMR lab call is open now, with a Teaming Partner List that connects private sector companies directly to national labs building commercialization project teams.

These efforts acknowledge what practitioners in the technology transfer space have long understood: the system requires a function that it does not natively produce.

What This Means for TTOs, Researchers, and Small Businesses The TRL 3 to 4 plateau is not a bug in the American innovation system. It is an emergent property of how research funding, institutional incentives, and procurement authorities are structured.

Technologies that stall at this stage often have strong technical merit, filed patents, and published validation data. What they lack is a defined buyer, field validation, and an operator committed to commercialization.

Addressing this gap requires inserting commercialization intelligence at the point where leverage is highest: at TRL 3 to 4, before institutional momentum dissipates and before the window for competitive commercialization closes.

That means evaluating market potential, competitive landscape, funding pathways, and dual-use opportunities at the disclosure stage, not after the patent issues. It means identifying the buyer, the market, and the transition pathway before the technology loses its champion or its funding runs out.

Before Bayh-Dole, no drugs were commercialized from federally funded university research. Since the act passed, over 153 FDA-approved drugs originated in university labs. From 1997 to 2007, university licensing created a $187 billion GDP impact and 279,000 jobs. UMN startups alone have attracted over $3.5 billion in IPOs, acquisitions, and funding.

The innovation system works. But only when someone bridges the gap between what the lab produces and what the market needs. The valley of death is not inevitable. It is a solvable problem. The question is whether you solve it before or after the window closes.

Sources Congressional Research Service, "Department of Defense Research, Development, Test, and Evaluation (RDT&E): Appropriations Structure," CRS Report R44711. Congressional Research Service, "Defense Primer: Research, Development, Test, and Evaluation," CRS In Focus IF10553, updated 2025. Congressional Research Service, "Small Business Research Programs: SBIR and STTR," CRS Report R43695. U.S. Government Accountability Office, "Directed Energy Weapons: DOD Should Focus on Transition Planning," GAO-23-105868, April 2023. U.S. Government Accountability Office, "Defense Innovation Unit: Actions Needed to Assess Progress and Further Enhance," GAO-25-106856, February 2025. Defense Innovation Board, "Terraforming the Valley of Death," Report on Strategic Capital, July 2023. National Science Foundation, Science and Engineering Indicators 2024, "Invention, Knowledge Transfer, and Innovation," February 2024. AUTM Licensing Activity Survey Data. University of Texas at Austin, Discovery to Impact 2025 Annual Report. University of Florida, FY2025 Technology Transfer Metrics. University of Minnesota, Technology Commercialization FY2025 Statistics.