DIB Innovators

Will Edwards, CEO & Co-founder of Firehawk walked around defense conferences with what was essentially Lego plastic demonstrating it could become rocket fuel. Everyone laughed until Firehawk proved thermoplastic works as a binder for 3D-printable solid propellant, cutting production from 60 days to 6 hours. The breakthrough came from a failed pivot: they tried selling hybrid rocket engines to disrupt the supply chain, but learned the military won't swap proven systems for new architectures. Success required replicating solid motors exactly, just manufactured differently. Will also explains why missile startups without propellant production will fail and why comparing defense manufacturing to SpaceX misses the point: primes already produce thousands of complex systems annually, they're just constrained by cast-and-cure physics, not capability. Topics discussed:Using thermoplastic binders to 3D print solid rocket propellant, cutting production time from 60 days to 6 hoursWhy hybrid rocket engine disruption failed and success required replicating solid motors with different manufacturing physics insteadHow traditional cast-and-cure propellant production constrains scale through 5,000-pound batches requiring hundreds of molds before curingWhy missile startups without propellant production capacity will fail competing against Northrop and L3Harris manufacturing queuesDebunking SpaceX comparisons in defense: primes already produce thousands of systems annually, constrained by physics not capabilityScaling from thermoplastic fuel experiments to 200,000 base bleeds annually and 10,000 rockets monthly by 2028Designing only for systems requiring 300,000+ units annually to ensure meaningful defense production impact at scaleBuilding distributed propellant manufacturing in Europe and Indo-Pacific regions to match Ukraine artillery consumption rates 

Zach Shore, President of Hermeus, and his team have demonstrated a turbine-based combined cycle engine in a wind tunnel for roughly $20 million. NASA and DARPA spent nine figures on the same architecture. The system uses proprietary modifications to an F100-229 to hit Mach 3, then routes airflow around the cocooned turbine directly into a ramjet to reach Mach 5. Reverse the process to decelerate. The result is reusable air-breathing flight from zero to Mach 5. No rockets required for acceleration or terminal glide on descent. They proved the complete architecture with a GE J85 engine. Now they're scaling to the F100-229 with the ramjet integration coming next.Quarterhorse, their 30,000-pound Mach 2+ aircraft powered by the F100-229, flies from White Sands this year. It's not a test article; it's the first platform with actual utility. Think unmanned F-16 capability stack: electronic warfare, rails for weapons, red air, high-speed target. Darkhorse will hit low Mach 5 before decade's end. Zach walks through why they're building metal airframes with removable panels and modular inlets rather than exotic composites, how they're using proven components like the F-16 landing gear and MiG fuselage design to avoid reinventing solved problems, and why demonstrating incremental capability beats PowerPoint pitches when you're trying to crack into integrated heavy systems programs. Topics discussed:Demonstrating turbine-based combined cycle engine architecture in wind tunnel for $20M versus NASA/DARPA's nine-figure development Routing airflow around cocooned F100-229 turbine into ramjet at Mach 3 to achieve reusable Mach 5 flight capabilityBuilding 10,000-pound unmanned aircraft from design to flight in 15 months to validate high-speed outer mold lineDeploying Quarterhorse 30,000-pound Mach 2+ platform with F-16 capability stack including electronic warfare and weapons railsDesigning metal airframes with removable panels and modular inlets rather than exotic composites for production scalabilityUsing proven components like F-16 landing gear and MiG fuselage design to avoid reinventing solved engineering problemsNavigating defense R&D contracting where labs compete with private innovators and prime contractors receive cost-plus incentivesBuilding hardware-rich iterative development approach with smaller engines before scaling to full F100-229 and ramjet integration  

Doodle Labs engineered multi-band radios that frequency-hop in milliseconds, outpacing jamming attacks that make satellite control impossible for real-time drone operations. When conflict in Ukraine started, the engineering team repurposed this architecture for anti-jamming by hopping between bands faster than adversaries could track and jam. Ashish Parikh, Co-CEO, details their monthly software release cadence driven by field engineers embedded with Ukrainian operators, where mission requirements emerged directly from combat: silent-mode mesh networking for operators who become targets the moment they emit RF, listen-only modes for monitoring video feeds without signature exposure, and coordinating multiple drones plus ground users while one expeditionary asset pushes 100+ kilometers into heavily jammed territory. Their modular architecture layers like Lego blocks that combine for complex mission profiles. Topics discussed:Repurposing multi-band capability from Army requirements into millisecond frequency-hopping anti-jamming during Ukraine conflictMonthly software releases driven by field engineers embedded with Ukrainian operators testing under active electronic warfareBuilding modular architecture layering sense anti-jamming algorithms, mesh optimization, and silent modes for coordination at 100km+ rangeDesigning point-to-point mesh networks for 20-100km missions requiring sub-100ms latency where satellite infrastructure adds prohibitive delayMaintaining NDAA-compliant supply chains with Singapore and US manufacturing capacity scaling to millions of units at commercial pricingDeploying listen-only modes and RF signature management so ground operators monitor feeds without becoming targetable network nodes

Oura tracked an armored unit's gunnery qualification and found baseline heart rate variability statistically predicted top-quartile performance on weapons platforms. Their three-tier privacy architecture gives individuals AI-driven feedback, lets them share data with coaches by explicit consent, and provides command aggregate de-identified metrics. This information helps command know not only when to back off on training, but when they can push harder and how trainees are recovering.Geoff Wylde, VP & General Manager of Health & Human Performance, discusses how Oura invested in NSA-approved hardware that operates in airplane mode for SCIF access and edge applications that sync intermittently with eight days of on-device storage for denied environments. Topics discussed:Tracking baseline heart rate variability to statistically predict top-quartile gunnery qualification performance on weapons platformsImplementing three-tier data architecture with individual insights, provider coaching access, and aggregate command-level readiness metricsAchieving 80% alcohol consumption reduction among special operators through biometric feedback loops from wearable ring dataBuilding consent-based privacy systems using de-identified accounts where no personal information enters company databases for DoDDeveloping NSA-approved hardware with airplane mode capability and Bluetooth antenna for SCIF access and classified environmentsCreating edge applications with 8-day on-device storage for intermittent sync in denied environments 

Clark Haymond, COO of DEWM, leveraged active duty relationships to compress validation timelines that would take years through traditional AFRL sequences. Having friends now in squadron leadership enabled direct jet time allocation for hypothesis testing, proving DEWM's DART could complete actual SEAD kill chains before committing to the formal program of record pathways. This network-driven approach delivers a combined operation with real-time feedback loops impossible in controlled SBIR phases.   Clark also explains why Air National Guard units became initial customers: state-controlled budgets enable faster procurement decisions than ACC. Guard wings need robust EW environments but face $5-15M price tags for traditional threat emitters. DEWM's 2-foot dish exploits the power threshold gap between legacy radar warning receivers and modern HTS pods, achieving tactically relevant detection ranges without massive amplifiers. Haymond candidly admits the MBA network delivered "almost none" of the value compared to Defense Ventures Fellowship placements and direct operator access.    Topics discussed:   Transitioning from permissive environments to building training systems for contested near-peer warfare scenarios Solving range emitter density problems where NTTR has tens of systems due to prohibitive replication costs 5th gen sensor sensitivity differences that enable lower-powered threat emitters to achieve tactically relevant detection ranges for training Building attritable radar targets using open architecture software-defined radios that customers can reprogram without vendor dependency or additional costs Leveraging active duty networks and squadron relationships to accelerate product validation cycles and compress AFRL timelines Comparing MBA network value versus Defense Ventures Fellowship placements for defense startup founders building hardware solutions  Scaling from Wessip weapon evaluation to international F-35 customers seeking cost-effective EW training infrastructure for SEAD mission preparation