Компания Tesseract Ventures выбрана для разработки автономной платформы управления дронами для ВМС США

Tesseract Ventures, a Kansas City-founded defense technology company, was chosen Tuesday by the Defense Innovation Unit to advance in one of the U.S. military's most consequential autonomous-systems programs — a competition to build a containerized platform that can store, launch, recover, and refit entire drone fleets with as few as two operators. The selection arrives two weeks after Defense Secretary Pete Hegseth signed a sweeping memo consolidating nearly all Pentagon drone programs under a single new office, making the Defense Innovation Unit the military's exclusive commercial interface for unmanned and autonomous systems and setting the stage for the first substantial production contracts to flow through exactly the kind of program Tesseract just entered. The company's entry in the competition is the NOMAD system — a mobile, containerized platform designed to manage large, heterogeneous drone fleets across land and sea using onboard intelligence and secure real-time data sharing. What separates NOMAD from conventional drone infrastructure is what it does between flights: the system is engineered to keep a fleet in a dormant readiness state, then autonomously launch, recover, and refit aircraft on command, without the one-to-one human attention that every current drone in the U.S. military's inventory still requires. What the CADDS Program Is Actually Solving The Defense Innovation Unit published its Containerized Autonomous Drone Delivery System solicitation in February 2026, framing the problem in blunt terms: "The Department of War faces a robotic mass challenge: current methods for deploying and sustaining unmanned aerial systems rely on direct human interaction to launch, recover, and refit each system. This 1:1 operator-to-aircraft model limits deployment speed and scale while exposing operators to unnecessary risks." That one-to-one ratio is the specific bottleneck CADDS is designed to break. In a contested environment — where the military needs to surge dozens or hundreds of drones simultaneously — the current model requires a proportional surge in trained operators. Personnel capacity becomes the ceiling on drone capacity. The program sought containerized systems that could collapse that ratio dramatically, with a stated preference for platforms operable by no more than two personnel handling all functions including storage, transport, launch, recovery, and refit. The requirements are technically demanding. The winning platform must support both homogeneous and heterogeneous drone mixes — meaning the container cannot be engineered around a single airframe, but must instead run a software layer that adapts to whatever mix of Government-directed UAS is loaded into it. Systems must be deployable from land or maritime platforms, operate in day and night conditions and inclement weather, and set up and tear down in minutes. Critically, the CADDS requirements include what the solicitation calls DDIL resilience: the ability to function in denied, degraded, intermittent, or limited communications environments — meaning the autonomy stack must sustain fleet management even when connectivity fails. All of this must be built on Modular Open Systems Approach (MOSA) principles, the DoD's legally mandated architecture standard that requires standardized, interchangeable interfaces and prevents proprietary lock-in. MOSA is not optional — it is the mechanism that ensures whatever container wins can later accept new drone types, new C2 software, and new payload configurations without a complete redesign. NOMAD's Software-Defined Architecture Tesseract's NOMAD platform addresses the heterogeneous-fleet requirement through what the company describes as a software-defined fleet management approach: an architecture that separates the fleet management logic from specific airframe characteristics, enabling the same container to adapt to different UAS types through software configuration rather than hardware modification. The system integrates four functional layers: open systems architecture, advanced autonomy software, resilient command and control, and scalable mission management. The resilient C2 component is the critical one for contested operations — it must maintain fleet coordination and mission continuity when the communications environment degrades, which is the baseline assumption in any high-intensity conflict scenario. The autonomy software handles the full operational cycle: dormant storage, launch sequencing, recovery guidance, and refit operations, all with what Tesseract describes as minimal operator cognitive workload. Tesseract announced the CADDS selection on July 14, 2026, with CEO and founder John Boucard noting: "Leveraging open systems architecture, advanced autonomy software, resilient command and control, and scalable mission management, Tesseract's solution increases operational flexibility while reducing operator cognitive workload." The existing market for containerized drone systems includes launchers from Rheinmetall, UVision, and others — but these platforms focus exclusively on the launch function. None currently integrates recovery and refit into the same containerized unit. That gap is precisely what CADDS is designed to close. A launcher that cannot recover its own aircraft forces operators to retrieve drones manually after each mission, which eliminates the autonomous lifecycle advantage and reintroduces the personnel dependency the program is trying to remove. What the Autonomy Switch Actually Controls The CADDS solicitation contains a requirement that deserves careful attention from anyone tracking how the U.S. military is approaching lethal autonomous systems: the platform must support both "operator-in-the-loop" and "operator-on-the-loop" decision-making processes. These are not synonyms. Operator-in-the-loop means a human being explicitly authorizes each significant action the system takes. Operator-on-the-loop means the system acts on its own while a human monitors and retains the ability to intervene. The distinction matters structurally: in a contested kinetic environment where a CADDS container is managing dozens of armed drones simultaneously, "on the loop" in practice means that individual strike authorizations may not occur. The system acts; the operator watches and can stop it; but the tempo of autonomous operations can outpace the bandwidth of human oversight. This is the live governance question at the intersection of fleet-scale drone autonomy and international humanitarian law — specifically the Principle of Distinction, which requires parties to a conflict to distinguish civilians from combatants before using force. A software system managing heterogeneous drone fleets across a dynamic battlefield at autonomous tempo occupies contested legal and ethical territory that international negotiations on lethal autonomous weapons systems (LAWS) have not yet resolved. The Pentagon's position, codified in DoD Directive 3000.09, rejects an international ban and instead maintains a governance framework requiring "appropriate levels of human judgment" — a formulation that leaves the line between the two loop modes operationally significant and institutionally ambiguous. None of this is unique to NOMAD or to Tesseract. It is the structural condition of the broader autonomous systems push CADDS represents. But the specific requirement to support both modes — written into the solicitation that Tesseract is now advancing in — confirms that the operator-on-the-loop model is an intended operational configuration, not a theoretical edge case. A Pathway to Fielding — and What It Is Worth Commercially The CADDS selection is not a final contract. Under DIU's Commercial Solutions Opening process, selected companies demonstrate their systems to validate operational performance. Demonstrations must be achievable within 90 days of formal contract award — meaning Tesseract's prototype testing is likely to begin in the fall of 2026. What comes after a successful demonstration is the commercially significant part. Under 10 U.S.C. § 4022(f), a competitively awarded prototype Other Transaction agreement that successfully completes can transition to a follow-on production contract without further competitive procedures. That follow-on production OT "will be available for use by one or more organizations in the Department of War," per the CADDS solicitation language, and could be "significantly larger than that of the prototype OT." In practice, a production contract won through the CSO/OT pathway is a potentially multi-service acquisition with no re-compete — a far more significant commercial outcome than the prototype funding itself. Other Transaction Authority contracts are also largely exempt from the Federal Acquisition Regulation and the Competition in Contracting Act, and GAO generally lacks jurisdiction to hear bid protests against OT awards. For commercial companies entering the defense market without traditional prime-contractor credentials, the OT pathway offers lower entry barriers and faster contract timelines — but with correspondingly reduced external oversight. Tesseract has navigated this landscape before. The company holds a prior OTA from U.S. Special Operations Command, a direct-to-Phase-II SBIR grant from the Space Force, and an Air Force contract at MacDill Air Force Base in Tampa — a track record that signals operational credibility to DIU evaluators even at the company's relatively small scale of approximately 20 employees and an estimated $10–25M revenue range. Inside the Pentagon's Autonomous Systems Reorganization The Tesseract selection lands inside a rapidly restructuring organizational context that directly determines what CADDS success is worth. In late June, Defense Secretary Hegseth signed a seven-page memorandum establishing the Direct Reporting Portfolio Manager for Unmanned Systems (DRPM-UxS) — a new office that absorbed the Defense Autonomous Warfare Group, Joint Interagency Task Force 401, and most other drone programs scattered across the military services, pulling them under a single director reporting to Deputy Defense Secretary Stephen Feinberg. The Pentagon named no interim director as of July 15. The memo's significance for commercial vendors is in one paragraph: "DIU will be the primary liaison with commercial companies for all unmanned and autonomous systems programs." A company that wins a CADDS production contract does not just sell to one service branch — it sells into a portfolio estimated at roughly $75 billion across programs, now managed through a single acquisition structure. The FY27 budget request included $54.6 billion specifically for the Defense Autonomous Warfare Group's drone development and procurement, described by Pentagon officials in the USNI-obtained memo as an unprecedented commitment. This restructuring follows years of fragmented, poorly coordinated drone procurement that blunted the Pentagon's first attempt at autonomous mass. The Replicator initiative — launched in August 2023 with a goal of fielding thousands of autonomous systems by August 2025 — delivered only hundreds by its deadline, according to the Congressional Research Service, hobbled by persistent technical failures, integration problems, and procurement software that could not command large numbers of heterogeneous drones simultaneously. T.S. Allen, Replicator's former director, said at a Brookings Institution event that his team helped supply "hundreds of drones to warfighters" but acknowledged the scale of need is far larger: "We probably need to 10x what Replicator delivered to meet the full demands of the modern battlefield." What Replicator revealed — and what CADDS is designed to address — is that the hardware gap is not the limiting factor. The limiting factor is the management infrastructure: the ability to autonomously store, cycle, and coordinate large numbers of drones in the field without proportional growth in human operators. That is the bet embedded in the CADDS program, and it is the bet Tesseract is now advancing in on. Is Anyone Else in This Race? Tesseract was selected alongside other unnamed companies in the CADDS competition — the DIU has not disclosed how many vendors advanced to the prototype phase. The solicitation attracted interest from companies already building drone-in-a-box systems for commercial applications, and defense industry analysts expected established players including Auterion's partners and defense primes with existing containerized launch concepts to participate. The program's recovery-and-refit requirement is its most technically differentiating aspect. Existing containerized launchers from Rheinmetall and UVision can deploy drones from a container, but neither handles post-mission recovery and refit within the same unit. CADDS requires the entire lifecycle — dormancy, launch, recovery, refit, re-launch — inside a single containerized footprint. Analysts noted in February 2026 that the closest existing system to the full CADDS capability was developed by a Chinese entertainment company for drone light shows. The Pentagon is asking American industry to replicate and militarize that capability. China's People's Liberation Army is not waiting for the U.S. to solve this. Chinese state media broadcast footage of the ATLAS drone swarm system in March 2026 — a ground-based platform built around the Swarm-2 ground combat vehicle. The competitive pressure that drove CADDS's creation has not slowed since the solicitation closed in February. Frequently Asked Questions What is the CADDS program, and what does Tesseract's selection mean? CADDS — the Containerized Autonomous Drone Delivery System — is a Defense Innovation Unit program seeking platforms that can store, launch, recover, and refit unmanned aircraft autonomously, with a crew of no more than two operators. The goal is to break the current one-to-one operator-to-drone ratio that limits how fast and how many drones the U.S. military can put in the air. Tesseract's selection means its NOMAD platform has passed DIU's initial screening and will participate in prototype demonstrations, expected to begin this fall. A successful demonstration opens a pathway to a no-competition production contract that could be awarded to multiple military services — a potentially significant commercial outcome. What makes the autonomous fleet manager technically harder to build than the drones themselves? The CADDS requirement that the container support "homogeneous and heterogeneous mixes of Government-directed UAS" is the central engineering challenge. A container engineered for a single drone type is a logistics upgrade. A container whose software can adapt to any mix of drones loaded into it — different sizes, different payloads, different recovery requirements — requires a software-defined fleet management architecture that separates control logic from airframe characteristics. Add the DDIL resilience requirement (the system must maintain autonomous operations even when communications are denied or degraded) and the MOSA mandate (all interfaces must be standardized and open for future upgrades), and the container's software is arguably more complex than the aircraft it manages. Does a drone fleet manager like CADDS change who authorizes the use of lethal force? This is the governance question the CADDS solicitation raises but does not answer. The program requires platforms to support both "operator-in-the-loop" (human authorizes each action) and "operator-on-the-loop" (system acts autonomously while a human monitors and can intervene) decision-making modes. In the latter mode, individual lethal engagements may not receive explicit human authorization — the system acts, the operator watches. Whether this satisfies the DoD's own Directive 3000.09 requirement for "appropriate levels of human judgment" in autonomous lethal systems, and whether it complies with international humanitarian law's Principle of Distinction, are questions that active diplomatic negotiations on lethal autonomous weapons have not resolved. The CADDS program does not claim to resolve them either — it is built for operational effectiveness in an environment where those questions remain open. How does the Other Transaction Authority pathway work for defense startups? DIU uses Other Transaction Authority — a legal instrument authorized under 10 U.S.C. §4022 and largely exempt from the Federal Acquisition Regulation — to award prototype contracts in 60 to 90 days rather than the year or more a traditional FAR-based procurement requires. If a prototype succeeds, the winning company can receive a follow-on production contract without further competitive procedures, available for use across multiple military organizations and potentially far larger than the prototype award itself. The trade-off: OT contracts carry reduced external oversight; GAO generally cannot hear protests against OT awards. For commercial companies like Tesseract that have not spent decades navigating the traditional defense industrial base, OTA is the primary available pathway into the Pentagon at commercial speed. ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.