Army SBIR in 2026: topic landscape and how to win

Army is the highest-volume DoD SBIR component

If you draw a map of DoD SBIR by topic count, Army covers the most ground. A typical cycle releases 80 to 120 Army topics, more than any other component. The volume is good news and bad news for a small firm. Good news because there are more chances to find a topic that fits your stack. Bad news because the topics are heterogeneous, the program offices behind them are many, and you cannot treat "Army SBIR" as a single buyer. You are actually dealing with ten to fifteen distinct buyers who happen to share a submission portal.

This post lays out the structure — who writes Army topics, how the 2026 cycles land on the calendar, how xTechSearch overlays with SBIR, and where an AI-first small business should actually compete. It is meant for a firm that can deliver code and models, not one that writes proposals for hire.

The program office structure behind Army topics

Army SBIR topics originate from three kinds of organizations: DEVCOM centers, PEO program offices, and Army Futures Command cross-functional teams. Each has its own culture, its own problem statements, and its own review style.

DEVCOM centers

DEVCOM (Combat Capabilities Development Command) is the engineering heart of Army research. It runs eight centers, several of which generate meaningful SBIR volume every cycle:

• DEVCOM C5ISR Center (Aberdeen, MD) — command, control, communications, computers, cyber, ISR. Heaviest AI buyer inside Army SBIR. Topics on sensor fusion, tactical networks, and cyber ML appear every cycle.
• DEVCOM Aviation and Missile Center (AvMC) (Redstone, AL) — aviation platforms, missile systems, air and missile defense. Predictive maintenance topics and autonomy topics cluster here.
• DEVCOM Ground Vehicle Systems Center (GVSC) (Warren, MI) — combat and tactical vehicles. Autonomy, condition-based maintenance, and digital engineering topics.
• DEVCOM Armaments Center (Picatinny, NJ) — weapons, munitions, fire control. More mechanical than AI, but fire control AI and counter-UAS topics appear.
• DEVCOM Analysis Center (DAC) (Aberdeen, MD) — modeling, simulation, operations research. Small but AI-heavy when it does post topics.
• DEVCOM Chemical Biological Center (Aberdeen, MD) — CBRN detection and protection. Overlaps with DTRA work.
• DEVCOM Soldier Center (Natick, MA) — individual soldier systems, textiles, nutrition. Smaller SBIR volume.
• DEVCOM Army Research Laboratory (ARL) — basic and applied research. ARL topics tend to be more speculative and research-heavy.

PEO program offices

Program Executive Offices manage acquisition programs of record. Their SBIR topics have something DEVCOM topics sometimes lack: a clear customer and a clear follow-on dollar. PEOs that generate AI-adjacent SBIR topics include PEO Intelligence, Electronic Warfare and Sensors (PEO IEW&S), PEO Aviation, PEO Ground Combat Systems, PEO Combat Support and Combat Service Support, and PEO Soldier. When a topic comes from a PEO, the transition path is more legible — you know which program of record is funding and who the government PM is.

Army Futures Command cross-functional teams

AFC runs the cross-functional teams (CFTs) that organize around specific modernization priorities: Long Range Precision Fires, Next Generation Combat Vehicle, Future Vertical Lift, Network, Air and Missile Defense, and Soldier Lethality. CFTs do not write SBIR topics directly very often, but the priorities they articulate shape what DEVCOM centers and PEOs ask for. Reading a CFT's published modernization priorities is the easiest way to predict which Army SBIR topics will be funded.

The 2026 Army SBIR calendar

Army participates in the DoD SBIR joint release schedule. In 2026 the Army-relevant releases are:

Release	Pre-release	Opens	Closes	Approx. Army topics
25.4 Release 12	Mid-April 2026	April 23, 2026	May 13, 2026	~50-70 Army topics (partial cycle)
26.1 Release 1	May 2026	May 7, 2026	June 3, 2026	~90-120 Army topics
26.2 Release 2	August 2026	Sept 2026	Oct 2026	~80-110 Army topics
26.3 Release 3	November 2026	Dec 2026	Jan 2027	~70-100 Army topics

Army also runs its own direct-to-Phase-II (D2P2) and Phase II Enhancement opportunities between releases. These are worth tracking if you have prior Phase I work, whether at Army or another agency. D2P2 requires documentation that the technology has reached the Phase II starting point through other funding — OSD, prior SBIR, or private.

xTechSearch: the prize-plus-introduction overlay

xTechSearch is the Army's prize competition program. It is not a contract, but it is a legitimate path into the Army ecosystem. The mechanics: a firm submits a short technical concept, progresses through pitch rounds, and wins cash prizes at each stage. The real value is not the prize money — it is the introductions. xTech finalists and winners get matched to Army program offices, briefed on relevant SBIR topics, and in some cases invited to direct-to-Phase-II opportunities.

For a firm with strong tech but no federal past performance, xTechSearch is a reasonable parallel track. The written concept is short, the time investment is manageable, and the downside is a few weeks of work. Firms that win xTech and convert to SBIR in the same fiscal year tend to have materially higher Phase I win rates on the topics they target, because the program office already knows their team.

Warm topic areas for AI firms in 2026

Based on prior cycle patterns, AFC CFT priorities, and the fiscal 2026 budget directives, the Army SBIR topic areas most open to a small AI firm without deep Army incumbency are:

Predictive maintenance and condition-based maintenance

AvMC, GVSC, and PEO Aviation post predictive maintenance topics every cycle. The Army has published data from its Integrated Data Environment (IDE) and the Logistics Modernization Program (LMP) that is usable for ML work once you have the right data access. Topics tend to focus on specific platforms: AH-64 Apache, UH-60 Black Hawk, Abrams M1A2, Stryker. A firm with prior ML-on-time-series work can compete credibly if it can articulate a clear path from historical maintenance records to deployable anomaly detection at the tactical edge.

Tactical logistics AI

The Army has an explicit modernization priority around contested logistics. Topics show up at DEVCOM Ground Vehicle, Army Sustainment Command, and the Logistics CFT. Real problems: demand forecasting at the forward support battalion level, last-tactical-mile routing under jamming, predictive parts sourcing. Data is uneven but there is a lot of it.

Intelligence processing and sensor fusion

PEO IEW&S, C5ISR Center, and INSCOM generate the highest-density AI topics. Problems include multi-INT fusion, signal processing for electronic warfare, SAR/EO/IR object detection, and natural language processing for OSINT. Classification is a gating factor — a firm without cleared personnel cannot work the classified variants but can still compete on unclassified enabling technology.

Autonomy and counter-autonomy

Ground and aerial autonomy topics appear every cycle. GVSC publishes Robotic Combat Vehicle related topics. AvMC publishes Future Vertical Lift autonomy topics. PEO Aviation publishes Launched Effects topics. Counter-UAS sits across Armaments Center and IEW&S.

Digital engineering and model-based systems

Less hyped than autonomy but steadier. Topics around digital twins, MBSE tooling, automated requirements checking, and simulation at scale appear across multiple DEVCOM centers. For a firm that can blend ML with engineering workflow tooling, this is underpriced.

Who wins Army SBIR topics, and why

Pattern recognition from the last five cycles of Army awards:

Practical steps for a first Army SBIR cycle

1. Read the topic titles for the prior two cycles. Patterns are visible. You can often tell which DEVCOM center will re-post a topic before the pre-release.
2. Pick two DEVCOM centers and one PEO. Not more. Depth in three buyers beats breadth across ten.
3. Contact the TPOC during pre-release. The Technical Point of Contact window is where legitimate conversations happen. Use it. Ask about the underlying problem, not about the proposal format.
4. Write to the problem, not to the topic. Topic language is often written to a specific unmet need. Reviewers reward proposals that name that need and show they understand the operational context.
5. Name the transition. Even a rough Phase III story — which unit, which program office, which follow-on vehicle — lifts a proposal out of the generic pile.
6. Submit on time with room to spare. DSIP portal issues in the last hour are a recurring cause of last-minute disqualifications.

The Army-specific failure modes

Failure modes that are more Army than anywhere else:

Bottom line

Army SBIR in 2026 is the largest single stream of federal AI topic volume. It rewards firms that treat it as a structured portfolio across DEVCOM, PEO, and AFC-priority topics, not a single lottery ticket. The warm areas — predictive maintenance, tactical logistics, intelligence processing, autonomy, digital engineering — each have a small set of incumbent firms and enough topic volume that a strong first-time firm can win one or two per cycle with focused effort. The firms that do well treat the program office as the customer, not the DSIP portal.

Frequently asked questions