The IDF views autonomy as a force multiplier that can save soldiers’ lives, accelerate decision‑making, and allow distributed units to operate with unprecedented precision.
By Hezy Laing
The IDF views autonomous systems as both a transformative opportunity and a strategic dilemma.
Senior officers in MAFAT, Unit 81, and the Robotics and Autonomous Systems Directorate see autonomy as a force multiplier that can save soldiers’ lives, accelerate decision‑making, and allow distributed units to operate with unprecedented precision.
Autonomous ground vehicles, drone swarms, and AI‑driven sensor networks promise faster target acquisition, reduced human exposure, and the ability to fight in multiple domains simultaneously.
At the same time, the IDF fears the risks that come with the atomization of its systems — the breaking apart of traditional platforms into semi‑independent nodes that sense, decide, and act with minimal human oversight.
Commanders worry about loss of control under electronic warfare, unpredictable swarm behavior, and the erosion of the human‑in‑the‑loop doctrine.
Here is a list of the IDF systems most likely to first become fully autonomous:
1. Roem / “Roem 360” autonomous drone swarm
Roem is built from the ground up as a networked swarm: dozens of small drones share data, self‑organize, and already perform semi‑autonomous recon and strike. That architecture makes it a natural candidate for early full autonomy, where human operators set mission goals and the swarm handles route planning, target selection, and coordinated attacks—especially against enemy drone swarms, artillery, and small units in dense urban terrain.
2. Iron Beam autonomous laser defense
Iron Beam’s engagement loop—detect, track, aim, fire—is already highly automated, because lasers must respond in fractions of a second to rockets, mortars, and drones. As autonomy deepens, it will likely run continuous, self‑optimizing defense over key bases and cities, automatically prioritizing threats, managing power loads, and coordinating with Iron Dome and other layers to create a largely self‑running “energy shield” around critical assets.
3. Carmel autonomous armored fighting vehicle
Carmel was designed as a “crew of one” vehicle with heavy AI support: autonomous driving, sensor fusion, and target recommendation. That same tech stack can evolve into a fully unmanned combat vehicle used for high‑risk urban spearhead missions, route clearance, and close support in areas saturated with anti‑tank weapons—sending robots first, humans later.
4. Jaguar autonomous border‑patrol robot
Jaguar already patrols border fences with autonomous navigation, cameras, and remote weapons, supervised from a control room. It’s likely to become fully autonomous in routine patrols, anomaly detection, and first contact: independently tracking infiltrators, marking routes, and even performing warning fire or non‑lethal actions before human commanders decide on escalation.
5. Harop / Harpy loitering‑munition family
Harop and Harpy already loiter, search for radar emissions, and dive autonomously on targets, with minimal human input. As autonomy expands, they’ll be used in large coordinated waves to suppress air defenses, hunt mobile command posts, and saturate enemy drone and radar networks—essentially acting as self‑directing “kamikaze clouds” over contested airspace.
6. SkyStriker loitering munition
SkyStriker is a precision loitering munition with advanced navigation and target‑tracking, making it easy to push more decision‑making onboard. In a fully autonomous mode, units could launch multiple SkyStrikers that independently patrol sectors, classify targets, and strike high‑value threats like ATGM teams, mortar crews, or moving vehicles, while humans set rules of engagement and no‑strike lists.
7. Fire Weaver autonomous fire‑control network
Fire Weaver already fuses sensors and shooters, recommending optimal weapons and firing solutions in real time. As autonomy deepens, it will likely manage entire battlespace fire plans: automatically assigning targets to tanks, artillery, drones, and infantry weapons, deconflicting friendly forces, and maintaining continuous “digital overwatch” that keeps lethal decisions fast but still under human veto.
8. Ghost & Mini‑Ghost micro‑UAVs
These micro‑UAVs are built for autonomous indoor and urban navigation, mapping rooms and alleys with minimal control. Full autonomy would see them used as swarms for building‑clearing recon, tunnel exploration, and close‑quarters surveillance—entering first, tagging threats, and feeding live 3D maps to troops who follow, or to other autonomous systems that act on that data.
9. Robattle unmanned ground combat platform
Robattle is a modular UGV designed for convoy protection, perimeter security, and urban combat, already capable of semi‑autonomous movement and sensor use. It’s poised to become fully autonomous in tasks like route clearance, decoy operations, and persistent perimeter defense—circling bases, reacting to gunfire or movement, and coordinating with drones and fixed sensors without constant human driving.
10. Blue Spear / Gabriel‑type autonomous anti‑ship logic
Modern anti‑ship missiles already fly autonomously, using waypoints, sea‑skimming profiles, and terminal seekers to find ships. As autonomy grows, naval doctrine may let these systems independently select the best target within a designated area, adapt routes to avoid defenses, and coordinate multi‑missile attacks—while human commanders define the engagement zone and strategic objectives, not each individual maneuver.
