The Army's launched-effects bet: a faster upgrade cycle for autonomous strike systems

The U.S. Army has identified three companies, AEVEX, Griffon and Dragoon, to provide its first long-range launched-effects capability, and it says production and delivery to Army formations are planned for 2026. The service is describing the systems up front as products of a Modular Open Systems Approach, or MOSA, built to improve situational awareness, precision targeting, interoperability and resilient kill chains (U.S. Army). In a separate announcement, the Program Manager for Uncrewed Aircraft Systems awarded AeroVironment a production contract for Long Range Reconnaissance Systems, a buy the Army intends to field through AeroVironment's P550 long-range reconnaissance drone (U.S. Army; FlightGlobal).

Read together, these are not two ordinary hardware purchases. They are declarations of a different acquisition philosophy. For decades, Army aviation and fires programs produced closed, single-prime platforms that took a decade or more to reach soldiers and were painfully slow to upgrade. The launched-effects strategy wagers the opposite: many smaller vendors, swappable modules, government-specified interfaces, and production rates measured in the thousands rather than the hundreds. The open question, the one that will determine whether 2026 becomes a turning point or a footnote, is whether that model can actually deliver autonomous systems in bulk, keep them working under electronic attack, certify their software fast enough to matter, and leave humans in control of the consequences.

What did the Army just buy?

The 2026 selections did not arrive out of nowhere. The launched-effects portfolio has been assembled in layers over several years, and the layering is itself the point.

In November 2023, the Army's Program Executive Office for Intelligence, Electronic Warfare and Sensors awarded other transaction agreements to Lockheed Martin and Northrop Grumman to mature infrared and electronic-warfare payloads for launched effects (DefenseScoop; Army Times). Northrop was awarded for two payloads and Lockheed for one; the initial awards were worth roughly $100,000 each, with the full three-phase effort expected to total around $37 million (DefenseScoop; InsideDefense). The plan was to take payloads from technology readiness level 6 to 7, integrate them into a captive-carry test event, and finish with integration and flight test on a chosen uncrewed vehicle, according to Dennis Teefy, the project director for sensors-aerial intelligence (InsideDefense). In March 2025, the Army tapped three companies for what it called a cutting-edge Launched Effects demonstration (U.S. Army), a stepping stone that service planners cited again in their 2026 manufacturing solicitation.

That sequencing reveals the architecture of the strategy. One layer of the portfolio develops payloads, the sensors and jammers, through organizations focused on electronic warfare. Another layer develops the air vehicles themselves, which is where AEVEX, Griffon and Dragoon now sit. A third layer, the AeroVironment P550 contract, buys a dedicated long-range reconnaissance platform through the uncrewed-systems program office. Rather than one contractor delivering one exquisite system that does everything, the Army is contracting for components of a kill chain and insisting that the interfaces between them belong to the government.

The schedule tells its own story. When the Army first described its launched-effects program of record in 2023, the first iteration was expected no earlier than fiscal year 2030, aimed at supporting the Future Attack Reconnaissance Aircraft and the Future Long-Range Assault Aircraft (InsideDefense). The Army canceled FARA in February 2024 as part of its aviation rebalance. Less than three years later, production and delivery of a long-range launched-effects capability to formations is planned for 2026. Either timeline could still slip, and the public announcements reviewed here do not disclose quantities or contract values for the new selections. But the direction of travel is unmistakable: pulled forward, not pushed back.

What exactly is a launched effect?

The Army's own definition is unglamorous. Launched Effects are uncrewed aircraft systems launched from a tube, either from air or ground platforms, and able to perform a variety of missions (Army SBIR|STTR Program). The family is defined by two design traits that carry most of the strategic weight. They are attritable or optionally recoverable, meaning the Army can accept losing them in combat at a price point where losing a crewed aircraft would be unthinkable. And they are built for mass, because the concept of operations imagines many of them employed together rather than a few flying alone.

What they carry depends on the layer. The 2023 awards focused on payloads for infrared sensing and electronic warfare: sensors that can see targets, and jammers that can blind an adversary, lofted forward of the main force (DefenseScoop). The 2026 long-range selections are framed around situational awareness and precision targeting feeding what the Army calls resilient kill chains (U.S. Army). The P550 sits at the reconnaissance end of the family. None of these systems is one thing, and that is deliberate. A tube-launched airframe is a bus; the mission is whatever module rides in it.

How modular design keeps vendors replaceable

Modular Open Systems Approach is the load-bearing phrase in the Army's announcements, and it deserves a plain explanation. Under MOSA, the government specifies the interfaces: the physical, electrical, data and software boundaries between a system's components. Any vendor that can meet those interface specifications can compete to supply a module, whether that module is a seeker, a radio, a jammer, a datalink or an entire airframe. The prime contractor becomes optional at the module level.

The standards work behind this is more mature than the rhetoric suggests. A 2026 research paper for the Naval Postgraduate School's Acquisition Research Symposium, written by Gregory Sanders of the Pentagon's Joint Production Accelerator Cell, catalogs the relevant architectures, including the Weapon Open System Architecture and weapons government reference architecture, the Open Mission Systems and Universal Command and Control Interface standard, the Future Airborne Capability Environment for avionics software, and the SOSA standard for sensor payloads. Army Launched Effects and the Long-Range Precision Munition program are among the paper's named case studies (Naval Postgraduate School).

Sanders's analysis, drawn from interviews with at least 18 current and former acquisition professionals, identifies several mechanisms by which open architectures expand production capacity. Open competition on severable modules lets the Army route around bottlenecks and rewards vendors for scaling quickly. Government reference architectures sustain competition between primes instead of accepting vendor lock. Commoditized modules lower barriers to entry for smaller companies, which eases surging in a crisis. Reuse of proven components across programs cuts development and test requirements, though the paper cautions that reuse demands strict conformity to the common standard (Naval Postgraduate School).

That last point is where software assurance lives. In a closed platform, assurance, meaning the verification that the software does what it should and nothing it should not, is a once-per-platform ordeal that can take years. In a modular system, the burden shifts: common software environments like FACE concentrate the certification effort on the interface standards and the verification process itself, so that swapping a module revalidates the module rather than the whole aircraft. Done well, this is how updates move at the pace of a war. Done badly, a multi-vendor stack multiplies the number of codebases, supply chains and test combinations an already stretched certification workforce must clear. The Army's public documents assert the architecture; they do not yet spell out how conformance and security will be verified across a shifting pool of competing suppliers. That gap is worth watching.

Can industry build airframes at $2,000 and 10,000 a month?

The production ambition embedded in this strategy is best illustrated by the Army's own manufacturing solicitation, released May 6, 2026 under SBIR solicitation 26.BX. It asks industry to demonstrate mass production of launched-effects airframes at an objective rate of 10,000 vehicles per month and an objective cost of $2,000 per assembled airframe, covering skins, stiffening elements, frames, control surfaces and other structural features. The manufacturing process itself must be modular enough to absorb a minor design change in hours or days (Army SBIR|STTR Program).

The solicitation is unusually candid about why current methods fail. High-performance carbon-fiber composites carry high material costs, long tooling lead times and labor-intensive fabrication that does not scale. Phase I offers up to $300,000 for a design and feasibility study; proposals were due June 24, 2026, a window that has now closed. Phase II demands a manufacturing demonstration of at least 80 airframes in one week, with one design change incorporated during the run (Army SBIR|STTR Program).

The distance between those two numbers is the honest measure of the challenge. Eighty articles in a week is a proof of method; ten thousand a month is an industrial base. The solicitation frames the first as evidence toward the second, not as the finish line.

The strategic backdrop, documented by Sanders, explains why the numbers are set so aggressively. The Army's munitions enterprise faces global proliferation of munitions, the battlefield arrival of attritable unmanned systems at scale, China's emergence as a manufacturing superpower, and what the paper calls the return of industrial-scale warfare. Against that, U.S. weapons production suffers from high barriers to entry, capacity fragmented program by program, and boom-bust contracting cycles that hollow out suppliers between conflicts (Naval Postgraduate School).

Brig. Gen. Rory Crooks, who leads the Army's Long-Range Precision Fires Cross Functional Team, described the same logic about launchers in early 2025. The service's in-house Autonomous Multi-domain Launcher, an uncrewed version of the HIMARS rocket launcher, which the Army live-fired at Yuma Proving Ground in April 2024, cannot be produced in useful numbers by the government alone. "That will be how we leverage industry," Crooks said of fielding autonomous launchers at scale. His three-pronged fires strategy pairs ammunition innovation with scaled launched effects and autonomy-augmented platforms, and the service has been weighing autonomous launchers that could fire larger weapons, including a proposed "Affordable High-Speed Strike" missile, earlier called the Precision Strike Missile Increment 5, that could be larger than today's 13-foot rocket pods (Breaking Defense). Launched effects are one moving part in a broader wager that affordability and openness, not technical edge alone, win industrial-age wars.

How do these systems survive electronic attack?

The electronic-warfare rationale for launched effects starts with an uncomfortable fact about emitters. Once a jamming capability switches on, it radiates. Anything that radiates can be sensed, geolocated and killed. The Army's own rationale for distributing electronic attack onto small, launched platforms is that an adversary who finds one finds a cheap drone, not a Terrestrial Layer System vehicle or the soldiers inside (DefenseScoop). Lifting sensing and jamming payloads into the air also extends their effective range past what ground platforms can reach, which is why the 2023 awards led with infrared and electronic-warfare payloads (DefenseScoop).

The 2026 selections extend that idea from payloads to platforms, with "resilient kill chains" named as an explicit design goal (U.S. Army). The mechanism the architecture enables is refresh speed. If a jammer's waveform is modular and the interface is open, an updated payload can come from any conforming vendor rather than from a single prime's upgrade cycle. That is the theory. Whether it holds depends on the conformance and software-assurance machinery described above, and on vendors actually investing in module improvements between contract bursts, something the boom-bust history of Army programs argues against taking for granted.

The funding picture shows the underlying work is treated as a continuing pipeline rather than a one-time fix. The service's fiscal 2026 budget request included $83.9 million for Electronic Warfare Advanced Technology (program element 0603275A), which matures electronic attack against an adversary's command and control, electronic support techniques that detect and geolocate hostile emissions to feed fires, and electronic protection for Army systems. That line is broader than launched effects and is not earmarked for them; it is the technology base from which launched-effects payloads draw (Fiscal Receipts).

Who keeps a hand on the trigger?

"Resilient kill chains" is the phrase that demands the human question. The 2026 announcements describe systems built for situational awareness and precision targeting, which are enabling functions, but launched effects as a family span from recoverable scouts to expendable munition-like aircraft, and the autonomy question scales with the mission.

The governing policy is Department of Defense Directive 3000.09 on autonomy in weapon systems, updated January 25, 2023. It requires that autonomous and semi-autonomous weapon systems be designed to allow commanders and operators to exercise appropriate levels of human judgment over the use of force, and it layers verification, testing and senior review requirements on top (U.S. Department of Defense). The phrase "appropriate levels" is deliberately elastic, and the elasticity is where most serious debate happens.

The Army's own testing agenda suggests it does not yet know what appropriate looks like in practice. Crooks, speaking about autonomous launchers and the pairing of crewed and uncrewed systems, described a live experiment rather than a settled doctrine. A crewed system has its own tasks, receive the mission, fire, resupply, and controlling autonomous systems simultaneously adds load nobody has yet measured. "It could be one-to-one," he said of the ratio of humans to autonomous platforms. "It could be one-to-N. We don't know what that is yet." The Army put those pairing questions to the test at Project Convergence Capstone 5, and Crooks framed the overall aim as making existing formations more lethal with autonomous systems rather than substituting robots for soldiers (Breaking Defense).

That framing matters because the industrial strategy and the human-control strategy are the same bet. A cheap, attritable, swappable fleet only exists if the command system can absorb it without drowning the humans responsible for its use.

Where the strategy can still break

The evidence assembled here supports the direction of the strategy, but it also documents its fragilities, and a few deserve plain statement.

The first is schedule churn. In late 2023 the official schedule put an initial launched-effects capability no earlier than fiscal 2030; the 2026 announcements plan deliveries within this calendar year. The compressibility of that timeline is the selling point of the new model, but it is also a reminder that Army schedules are political artifacts. Breaking Defense noted in February 2025 that priorities could shift with new civilian leadership, and the Army had already shown, by canceling FARA a year earlier, that anchor programs are not sacred (InsideDefense; Breaking Defense).

The second is the manufacturing gap. Ten thousand airframes a month at $2,000 each are stated objectives, not demonstrated capabilities, and the accepted proof point, 80 airframes in a week, is roughly a thirtieth of the weekly pace the goal implies (Army SBIR|STTR Program). The third is economic. Sanders's analysis warns that optimizing module competition for price can destroy the slack capacity that makes surging possible, the same trap as any just-in-time supply chain asked to behave like a war footing (Naval Postgraduate School).

The fourth is the unwritten assurance regime. MOSA shifts software risk from one big prime to a mesh of small vendors, and public materials do not yet explain how conformance, cybersecurity and verification will be enforced across that mesh at speed. Finally, nothing in the sources guarantees resilience against an adversary whose own jammers and adaptation cycles keep pace. Electronic superiority claimed at fielding decays; the architecture's promise is that the payload, and hence the countermeasure update, can be swapped faster than the threat moves. That promise is currently an assertion, not a demonstration.

What success would prove

The near-term tests are concrete enough to check. Deliveries to formations planned for 2026 either arrive or slip. Phase II manufacturing demonstrations either hit 80 airframes in a week with a mid-run design change, or they do not. Modules from one vendor either swap into another vendor's airframe without a bespoke integration effort, or they do not. Each result will say more about the future of defense acquisition than any policy memo has since the Army committed to open systems.

The deeper significance is the inversion it proposes. For most of the precision-strike era, the United States bet on a small number of exquisite platforms per contractor per decade. The launched-effects strategy bets that a government-owned architecture, enforced standards and a standing pool of competing vendors can turn weapons production into something closer to a continuously refreshed supply chain, resilient because it is distributed, current because its parts are replaceable, and survivable because no single loss, of a platform, a factory or a jamming waveform, breaks the chain. As of September 2026, the Army has chosen that experiment on the record and put three relatively small companies and AeroVironment on the calendar to prove it. Whether industry can build the airframes, whether the software regime can keep pace, and whether commanders can hold the judgment the directive requires will be answered by hardware delivered, or not delivered, in the months just ahead.

Sources

• U.S. Army, "Army Accelerates Long Range Launched Effects", army.mil
• U.S. Army, "PM UAS Awards Production Contract for Long Range Reconnaissance Systems to AV", army.mil
• FlightGlobal, "US Army to field AV P550 long-range reconnaissance UAS", March 2026
• U.S. Army, "U.S. Army Taps Three Companies for Cutting-Edge Launched Effects Demonstration", army.mil, March 27, 2025
• Ashley Roque, "Eyeing a new autonomous launcher prototyping initiative, Army checking out industry options", Breaking Defense, February 4, 2025
• Mark Pomerleau, "Lockheed Martin, Northrop Grumman score awards for Army's 'launched effects' initiative", DefenseScoop, November 29, 2023
• Colin Demarest, "Lockheed, Northrop win US Army contract for spy gear on launched drones", Army Times, November 29, 2023
• InsideDefense, "Army awards OTAs to Lockheed Martin, Northrop Grumman for Launched Effects program", November 2023
• Army SBIR|STTR Program, "Scalable Agile Manufacturing of Launched Effects", solicitation 26.BX, released May 6, 2026
• Gregory Sanders, "Architecting Affordable Mass", Joint Production Accelerator Cell, Acquisition Research Symposium, Naval Postgraduate School, May 7, 2026
• Fiscal Receipts, "Electronic Warfare Advanced Technology: FY2026 Budget" (program element 0603275A)
• U.S. Department of Defense, "DoD Directive 3000.09: Autonomy in Weapon Systems", updated January 25, 2023