DARPA's May 6 Drop: Four Biotech and Photonics SBIR Topics Closing June 3 Reshape the Defense Innovation Pipeline

DARPA's Biological Technologies Office and Microsystems Technology Office quietly pre-released four FY26 SBIR topics on April 30 with a topic open date of May 6 and a hard close at the end of business on June 3. The window is short by design. DARPA has moved over the last three years from large batch BAAs released twice a year to a continuous monthly drip of small, sharply-scoped topics that close roughly four weeks after they open. For small businesses with the relevant technology already on the shelf, the new cadence is a clear advantage — fewer competitors, faster decisions, and a topic that almost certainly maps to specific research portfolios DARPA program managers are trying to fill right now. For small businesses that need three months to assemble a serious proposal, the new cadence is a problem.

The four May 6 topics — Smart Whole Blood Field Transfusion (SWiFT), Broadening Availability of Regimens for K-9s (BARK), Extremity Platform for On-demand Surgical Implantation (EXPOSITION), and Photonic-electronic Panels Integration (PEPI) — sit at the unusual intersection where defense-mission urgency, late-stage biomedical engineering, and advanced photonics meet. Three of the four are inside the Biological Technologies Office, which has been the most active SBIR shop at DARPA across FY25 and FY26. The fourth is a Microsystems topic. All four are small-business eligible at the Phase I gate, and three of the four offer a Direct-to-Phase-II pathway for firms with sufficient prior-art evidence to skip the feasibility study.

SWiFT: a $300K-to-$1.8M problem worth solving

The Smart Whole Blood Field Transfusion System (topic code DPA26BZ01-DV003) asks small businesses to design a self-contained, autonomous device that can collect blood from a donor in the field, store it temporarily for less than four hours, run diagnostic testing for blood type and infectious agents, and execute the transfusion — all without requiring a deployed clinical team. Phase I awards are $300,000. Phase II awards are $1,800,000. Firms with a sufficiently mature prototype can apply directly for the $1,500,000 Direct-to-Phase-II path. The combination of Phase I plus full Phase II ceiling brings total non-dilutive funding for a successful awardee above $2 million.

The reason SWiFT is the strongest of the four topics for civilian medical-device firms to consider is that the dual-use angle is enormous. The same self-contained field transfusion technology that could keep wounded troops alive in austere environments has direct civilian application in rural emergency medicine, disaster response, and pre-hospital trauma care. A Phase II SWiFT prototype with FDA clearance pathway design embedded into the engineering would be an immediate acquisition target for the major medical-device manufacturers — Stryker, Medtronic, and Becton Dickinson all maintain active scouting programs for exactly this kind of field-deployable trauma technology. The combination of non-dilutive DARPA funding plus a clearly defined civilian exit is exactly the profile that distinguishes the SBIR programs from generic venture capital.

BARK: K-9 therapeutics as a backdoor into translational medicine

The BARK topic (DPA26BZ01-NP001) is an open-topic SBIR with a much smaller Phase I ceiling — $150,000 — and a Phase II ceiling of $1,300,000. It asks for medical products that are interoperable across humans and dogs. The headline framing is military working dog medicine: the Department of Defense maintains a working dog population of approximately 2,500 animals across all services, and the operational expectation that a battlefield medic can use the same therapeutics on a wounded handler and on the dog at the handler's side is a real combat-medicine problem.

The deeper read is that BARK is a translational medicine accelerator dressed in defense clothing. Veterinary therapeutic development is regulated by a different FDA center, moves faster, and has lower clinical-trial costs than human therapeutic development. A small business that can develop a therapeutic in the K-9 channel under BARK funding, demonstrate safety and efficacy, and then carry the same molecule into human Phase I trials has a substantially shortened path to a human therapeutic. Several biotech firms have used variations of this strategy to develop oncology drugs by running canine trials first, where the cancer biology is closer to humans than rodent models are. BARK is, in effect, a DARPA-funded version of that strategy. Phase I awards are modest, but the strategic value of the canine-to-human pathway is much larger than $150K suggests.

EXPOSITION: finger regeneration as a Direct-to-Phase-II topic

EXPOSITION (DPA26BZ01-DV004) is the most technically ambitious of the four. It asks for an on-demand regenerative medicine platform that can produce a complete restored finger — bone, cartilage, soft tissue, nerve — following traumatic amputation. The topic is structured as a Direct-to-Phase-II-only award at $1,500,000. There is no Phase I gate. The signal is that DARPA only wants firms with substantial prior work in tissue engineering, decellularized scaffolds, or 3D bioprinting of complex composite tissues. A first-time tissue-engineering firm with no relevant publications or prototypes will not be competitive.

For the firms that are competitive — likely a population of fewer than two dozen U.S. companies in the bioprinting and regenerative medicine space — EXPOSITION is the most consequential single Direct-to-Phase-II topic of FY26 in this domain. The narrow scope (finger restoration specifically, not generic limb regeneration) is a calculated DARPA program-management move. By constraining the problem to a single anatomical structure with a defined surgical insertion pathway, the program manager has made the milestone definition tractable inside an 18-to-24-month Phase II clock. Firms with the relevant prior art should treat EXPOSITION as a near-mandatory submission, and firms without it should not waste cycles on a proposal that will not score.

PEPI: photonic-electronic integration at chip-scale

The fourth May 6 topic, Photonic-electronic Panels Integration (DPA26BZ01-NV002), is from DARPA's Microsystems Technology Office rather than BTO. It asks for prototype technology that supports integration of photonic chips with high-density three-dimensional chip-to-chip and intra-chip optical routing. The application is data-center-scale and AI-compute-scale photonic interconnect — exactly the domain that NVIDIA, Intel, and Broadcom are spending billions to solve through their own R&D budgets.

PEPI matters for SBIR strategy because it is one of the relatively rare Microsystems-office topics that opens directly to small businesses without requiring a prime-contractor partnership at submission. For startups in the silicon photonics space — the broader category that includes Ayar Labs, Lightmatter, and several venture-backed firms now competing in optical compute — a DARPA SBIR award at Phase I or Phase II would be a meaningful credential in the same fundraising cycle. The $300K Phase I is small relative to typical Series A check sizes in this category, but the validation effect is substantial.

What the cadence change means for small-business strategy

DARPA's shift to monthly topic releases with four-week response windows changes what a serious SBIR strategy looks like. The old model — twice-yearly large BAAs with two-to-three-month response windows — rewarded firms that did most of their proposal work after the topic dropped. The new model rewards firms that maintain a continuous library of pre-drafted technical narratives, with management and budget templates already polished, ready to be fitted to a topic that drops on the first Wednesday of any given month. The cost of operating in the new cadence is higher business-development overhead. The benefit is that a firm with a well-maintained library can submit to three or four DARPA topics per year instead of one, materially raising the expected value of the SBIR pipeline.

Firms that have not built that library are at a structural disadvantage in the new cadence, and the disadvantage compounds. DARPA program managers see the same firms submitting to multiple topics across multiple months, build relationships with those firms across submissions, and develop intuitions about which firms can execute. The firms that show up only when a topic happens to align with their core business get treated as one-off submitters and rarely advance past the Phase I gate. The continuous-engagement model is now the dominant model for serious DARPA SBIR strategy.

The deadline reality

Proposals for all four May 6 topics are due June 3, 2026, through the DoD SBIR/STTR Innovation Portal at dodsbirsttr.mil. The clock is real — DARPA does not extend SBIR deadlines under normal circumstances. Firms that have not started a proposal by the time this is published have approximately two and a half weeks to draft, internally review, and submit a competitive package. For SWiFT and EXPOSITION, the prior-art demands make this difficult for first-time applicants. For BARK and PEPI, the technical scope is broader and the entry barrier is lower, but the four-week-from-topic-open window still favors firms with operational SBIR infrastructure in place.

Our companion analysis of the recent NASA SBIR/STTR FY26 BAA shifts covers the broader 2026 SBIR landscape across agencies. For firms running a multi-agency SBIR strategy, DARPA's monthly cadence and NASA's annual cycle now require different operational models — and the firms winning the most non-dilutive dollars in 2026 are running both in parallel.