What is Manned-Unmanned Teaming (MUM-T)? AI Autonomy Explained

Without This Technology, Next-Generation Capabilities Are Grounded

Imagine an F-35 pilot in a high-threat environment, already managing their own aircraft, sensors, and defensive systems. Now, tell them to also fly four "loyal wingman" drones simultaneously. This is a human impossibility; the pilot's cognitive load would instantly exceed 100%, leading to catastrophic failure. This is the central bottleneck of future air combat.

Manned-Unmanned Teaming (MUM-T) is the revolutionary concept designed to solve this. Its core is not "remote control"; it's an "AI Autonomy Core"—a trusted "AI co-pilot" or "flight lead" installed on the unmanned aircraft. The human pilot no longer issues micro-commands like "fly left, turn on sensor." Instead, they issue "mission-level intents" like "scout Area B" or "suppress that SAM site." The AI wingman then autonomously plans its own route, avoids threats, and executes the task. Without this AI core, multi-billion-dollar 6th-generation fighter programs (like NGAD) cannot execute their "distributed lethality" doctrine, and their "Collaborative Combat Aircraft" (CCA) would be nothing but expensive, vulnerable, remote-controlled toys.

The Core Technology Explained: Principles and Generational Hurdles

Past Bottlenecks: Why Legacy Architectures Failed

The legacy model of unmanned warfare, epitomized by the MQ-9 Reaper, is a "one-to-one" or "many-to-one" control ratio. A large team of pilots, sensor operators, and analysts in a ground station thousands of miles away is required to "remote pilot" a single drone. This architecture has three fatal flaws for a peer conflict:

1. Massive Manpower Sink: It requires more personnel to operate one unmanned Reaper than to fly one crewed F-35.

2. Fragile, High-Bandwidth Datalinks: It relies on high-bandwidth, beyond-line-of-sight satellite communications, which are a prime target for jamming and attack in a contested environment.

3. Unacceptable Latency: The signal delay (latency) is too high for the split-second decisions required in a dynamic, high-threat battlespace.

This "remote piloting" model is completely unworkable for the close, real-time coordination demanded by MUM-T.

What Is the Core Principle?

The core principle of MUM-T is a fundamental shift from Remote Piloting to Mission Command. It is a revolution in the human-machine relationship, defined by "delegated authority" to the AI.

This Autonomy Core operates on a logic loop:

1. Sense & Fuse: The AI wingman uses its own multi-domain sensors (radar, EO/IR, RF) to build a 360-degree picture. It also fuses data from the crewed lead aircraft and other off-board sources into a single, unified operational picture.

2. Understand & Decide: This is the leap. The AI doesn't just "see" a target; it "understands" the tactical situation. When the human pilot issues the intent "screen my flank," the AI autonomously calculates: "To screen, I must hold this orbit, use this sensor mode, and position myself between my human lead and this known threat axis."

3. Act & Team: The AI autonomously executes flight control, sensor management, and (when authorized) weapon employment. If Wingman 1's sensor is jammed, it autonomously communicates with Wingman 2 to take over the track, with no human intervention required.

4. Report & Adapt: The AI only communicates "decision-level" information back to the human, e.g., "Threat identified, request authorization to engage" or "Bingo fuel, RTB." This "curation" of data is what mitigates the mission risk of overloading the pilot.

The fundamental design goal is to elevate the human from an "operator" to a "commander," allowing them to focus on high-level strategy while the AI handles the complex, high-workload tasks of execution.

Breakthroughs of the New Generation

• Scalable Command: It allows one human to command a "mass" of 2, 4, or 8 unmanned assets, exponentially increasing the force's combat power.

• Risk Mitigation (Expendable Mass): The unmanned platforms are "attritable" (cheaper, expendable). They can be sent on the most dangerous tasks—flying ahead to find SAMs, penetrating A2/AD bubbles, or conducting electronic attack—while the $100M+ crewed platform stays at a safe standoff distance.

• Tactical Flexibility: The AI wingmen can act as "forward sensors" (allowing the F-35 to remain stealthy) or "external weapons magazines" (allowing the F-35 to cue targets for the wingman to shoot), creating a multitude of new offensive and defensive plays.

Industry Impact and Applications

The Implementation Blueprint: Challenges from Lab to Field

Making MUM-T a combat reality requires surmounting three immense systems engineering challenges: AI trust, human-machine interface, and open standards.

Challenge 1: The AI "Trust" Problem (Verification & Validation)

In the chaos of combat, how do you prove the AI won't make a catastrophic mistake? How do you certify an AI for combat? This is the central risk mitigation challenge.

• Core Tools and Technical Requirements:

  • High-Fidelity Synthetic Environments: This is the AI's "dojo." Using Digital Twins of the platforms and the threat environment, programs (like DARPA's ACE) must run the AI through millions of hours of virtual, high-threat scenarios to validate its behavior and build the "trust" required for deployment.

  • Explainable AI (XAI): The AI's decision-making process cannot be a "black box." It must be able to communicate why it is making a recommendation.

Challenge 2: The HMI & Cognitive Load Problem

The success of MUM-T hinges on the cockpit. If the HMI is confusing or overwhelming, the entire concept fails.

• Core Tools and Technical requirements:

  • AI Mission Manager: This is the software "co-pilot" that acts as the human's "translator." It must curate the tidal wave of data, filter out the noise, and present the pilot with simple, clear, tactical options, not raw data streams.

  • Ruggedized Edge Compute: This AI software runs on powerful, "tactical edge" computers. These units must be hardened to survive the G-forces, temperatures, and vibrations of a fighter cockpit, all while running complex AI inference models in real-time.

Challenge 3: The Interoperability & Open Architecture Problem

How do you ensure a Lockheed Martin F-35 (a U.S. asset) can command a BAE Systems drone (a U.K. asset) in a future NATO operation?

• Core Tools and Technical Requirements:

  • Modular Open Systems Approach (MOSA): A U.S. DoD mandate that forces contractors to use standardized, non-proprietary interfaces for hardware and software.

  • Universal Command and Control Interface (UCI): A government-owned standard for "speaking" to unmanned systems. Adopting MOSA and UCI is the only way to enable interoperability, prevent vendor lock-in, and allow for the rapid "plug-and-play" insertion of new autonomous capabilities.

Kingmaker of Capabilities: Where is This Technology Indispensable?

MUM-T is the new operational doctrine for all services:

• Air Combat: The USAF's Collaborative Combat Aircraft (CCA) program, the core of the Next Generation Air Dominance (NGAD) family of systems.

• Army Aviation: Future Vertical Lift (FVL) platforms are being designed from the ground up to command and control "Air-Launched Effects" (ALE) drones.

• Ground Combat: Robotic Combat Vehicles (RCVs) scouting ahead of and protecting crewed tanks like the Abrams.

• Naval Warfare: Unmanned Surface Vessels (USVs), like those in Task Force 59, acting as distributed sensors and shooters for crewed destroyers.

The Road Ahead: Human-on-the-Loop

The immediate challenge is building trust in the AI and hardening the datalinks that connect the team. The next evolution is the "human-on-the-loop" model: the AI swarm is given a mission and executes it fully autonomously, only "asking" for human authorization at the final "weapons release" step, further accelerating capability deployment in contested environments.

The Investment Angle: Why Selling Shovels in a Gold Rush Pays Off

The pivot to MUM-T and "attritable" aircraft has ignited the largest R&D and acquisition boom since the dawn of stealth. In this revolution, the true, high-margin value is not just in the "bodies" (the drone airframes), but in the "brains" and "nervous system" that make them work.

The "shovels" in this new gold rush are the enabling technologies:

1. The Autonomy Core Software: The AI "brain" itself. This is the intellectual property crown jewel for primes like Northrop Grumman, Lockheed, and BAE Systems.

2. The HMI & Edge Compute Hardware: The rugged, AI-capable mission computers and displays that live in the cockpit and on the unmanned platform.

3. The Synthetic Training Environment (STE): The digital twin and simulation software used to "build, test, and validate" the AI.

Compared to betting on which single airframe will win a contract, investing in these "platform-agnostic" enablers—the AI brains, the HMI, and the digital validators—provides broad, durable exposure to a fundamental, multi-decade transformation in warfare.

Aminext is a small blog, if you found this article insightful, would you consider sharing it with others or giving it a "like"? Every little bit of support is a huge encouragement for me to keep tracking these trends for you. Thank you so much for reading!