Click to see the full-size image
Cheap drones, heavy bombs, disrupted shipping, and vulnerable supply routes are challenging the idea that technological superiority guarantees military success. The emerging contest is not simply over who builds the smartest weapon. It is over who can adapt, produce, and keep fighting after the enemy learns how to break it. Washington’s Project Meridian offers one answer. Recent battlefields suggest a much harder test.
The Battlefield Rewrites the Blueprint
The war in Ukraine shows that military innovation does not always mean more sophisticated weapons. Sometimes it means finding a simple way around an enemy’s defenses. Russian forces did that with drones guided through fiber-optic cables instead of radio signals, allowing them to operate despite radio jamming.
During the fighting to push Ukrainian troops out of Russia’s Kursk region, fiber-optic drones threatened the roads supplying Ukrainian positions. Sustained attacks on transport made those positions harder to supply, reinforce, and evacuate. Ground that remained defensible against a direct assault became increasingly difficult to hold as supply routes came under pressure.
Drones did not replace infantry or artillery. Their value came from linking reconnaissance and strikes to a broader effort to disrupt Ukrainian logistics. A relatively small change in drone guidance helped Russian forces put pressure on the supply lines sustaining the defense, rather than just individual vehicles or positions.
A separate example from eastern Ukraine shows how drone attacks can also threaten robotic supply vehicles
Nor did the advantage remain exclusive. Ukrainian units expanded their own fiber-optic capabilities, turning a Russian breakthrough into another round of adaptation. Cable guidance removes the vulnerable radio-control link, but introduces weight, handling, and physical constraints. It does not make the aircraft invisible or indestructible. A solution to one problem creates a different set of tradeoffs.
Russia’s glide bombs demonstrate another route to military innovation. Instead of replacing existing weapons entirely, guidance and wing kits give older bombs greater reach and precision. Their heavy payloads perform tasks that small drones cannot. Their use against Ukrainian positions illustrates why the future battlefield will contain both cheap electronics and large quantities of conventional explosives.
The broader lesson is uncomfortable for anyone selling a universal replacement for the old military. New systems do not arrive on an empty battlefield. They must work with existing aircraft, artillery, engineers, supply units, and command structures. Their value depends on how those parts fit together.
That also limits what strike videos can prove. A successful attack demonstrates an outcome, not the average success rate, the number of failed missions, or the effort required to repeat it. The decisive question is whether an army can turn occasional success into a dependable military capability.
The Cost of Keeping Routes Open
The Red Sea offers a different warning. An attacker does not necessarily have to defeat a navy to impose strategic costs.
Houthi attacks pushed shipping companies toward longer routes around Africa. By late summer 2026, traffic through the Suez Canal was recovering but remained well below pre-crisis levels. Houthi attacks had pushed shipping companies onto longer, more expensive routes, showing how a regional armed group could disrupt global trade without defeating a major navy.
An interception can therefore be a tactical success without restoring normal commercial activity. The defender must protect ships and sustain confidence over time. The attacker may only need to keep the risk credible. Counting destroyed drones does not answer whether the shipping route is functioning.
The familiar comparison between a cheap drone and an expensive interceptor captures only part of this problem. Saving a merchant ship or warship can justify an expensive defensive shot. The deeper question is whether sufficient interceptors, maintenance capacity, crews, and replacement stocks remain available throughout a prolonged campaign.
Ukraine presents a related problem on land. Russian glide bombs combine substantial destructive power with relatively inexpensive conversion kits. Defending against every incoming weapon with a costly missile is not necessarily sustainable. But abandoning high-end air defenses would leave other threats unanswered.
The likely response is a layered force rather than a single miracle weapon. Different threats demand different combinations of detection, electronic warfare, physical protection, and interception. An army also needs to keep operating when some threats get through.
Kursk and the Red Sea are not interchangeable battlefields. One concerns a contested land campaign, the other pressure on maritime commerce. What connects them is the importance of maintaining a function under repeated attack. Delivering supplies and keeping shipping moving matter more than producing an impressive interception tally.
Washington Looks Upward
Washington is approaching this problem through a different starting point: global networks, industrial scale, and institutional control.
On September 30, Pete Hegseth announced Project Meridian alongside a proposed Autonomous Warfare Command and the Fortress America initiative. Meridian would examine future military capabilities, while the new command would organize the expansion of autonomous and robotic systems. Fortress America would address the resilience of bases and supply chains. These are related initiatives, not a single approved war plan.
U.S. service members at the State of the Force address, Quantico, September 30, 2026
Meridian’s advisers include Elon Musk, Palmer Luckey, and Newt Gingrich, under Pentagon technology chief Emil Michael. The study has 120 days to report. Its significance is not that three men can invent the next war. It is that major technology suppliers are being invited to help define the military capabilities Washington should pursue.
The September 30 memorandum commissions a study of future warfare
The State of the Force event at Quantico, where Project Meridian was announced, September 30, 2026
Some of the supporting infrastructure already has funding. On May 26, Space Systems Command awarded SpaceX a $2.29 billion agreement for the Space Data Network Backbone. A fully operational prototype is due by the end of 2027. The intended function is to connect military sensors and weapons through a high-capacity orbital communications network.
That addresses a real military requirement. Information has limited value if it cannot reach the people or systems that need it. But a larger network does not automatically become a more dependable one. The test is whether useful connections survive disruption, and whether units can continue operating when they do not.
Heavy-lift rockets could support the deployment of larger satellite constellations, although these programs do not yet amount to a single publicly defined architecture. Starship Flight 14 reached orbit on September 28 and deployed 26 Starlink V3 satellites. That demonstrated payload delivery, not every capability promised for the rocket. Reaching orbit once and providing dependable, economical military infrastructure are different milestones.
Starship Flight 14 launches from Starbase, September 28, 2026, carrying 26 Starlink V3 satellites
Starship’s final descent, landing burn, and splashdown during Flight 14, September 28, 2026
Golden Dome extends the ambition to missile defense. In May, the Congressional Budget Office estimated that a notional system consistent with the administration’s stated objectives could cost about $1.2 trillion over 20 years. Crucially, this was not a price tag for the Pentagon’s approved design. The final architecture had not been publicly disclosed.
A notional national missile-defense architecture assessed by the Congressional Budget Office
Estimated costs of the Congressional Budget Office’s hypothetical system, expressed in 2026 dollars
The policy risk is not simply excessive ambition. It is defining military problems in ways that favor the products already offered by influential suppliers. That deserves scrutiny without assuming that commercial involvement makes every proposal useless. A contract is evidence of spending. It is not evidence of battlefield superiority.
Elon Musk, named a co-leader of Project Meridian
Machines Still Need Factories
The American approach emphasizes integrating ambitious systems through large institutions. The battlefield experience points toward an equally important requirement: changing ordinary systems quickly.
These approaches are not mutually exclusive. Satellites and inexpensive drones can support the same force. The tension lies in their development cycles. A major space program cannot be redesigned every few weeks. A frontline drone or its communications equipment may lose effectiveness if updates take years.
The industrial base must therefore support both continuity and rapid revision. Ukraine’s drone sector demonstrates the importance of close feedback between operators and developers, but also the vulnerability of assembly dependent on imported components. A domestically assembled aircraft is not necessarily an independently sustainable capability.
Russian developers are also discussing the problem in terms broader than individual weapons. In September, the Ushkuynik center presented a drone assembly line housed in a modified shipping container. The proposal illustrates an effort to distribute production rather than concentrate it in a single factory.
These are developer proposals, not independently demonstrated solutions at national scale. Nevertheless, they identify two important problems: distributing production and closing gaps between separate defensive systems.
A workshop, a repair team, and a reliable component supplier may therefore matter as much as another model advertised as revolutionary. Military adaptation requires people who can identify a failure, explain it to engineers, approve a modification, and return workable equipment to service.
Autonomous Warfare Command: new drones, familiar geometry. All that’s missing is the secret handshake
Autonomy does not remove these obligations. The meaningful question is which task a machine can perform reliably without continuous human input, under what conditions, and with what consequences when it fails. “Autonomous” should describe a tested function, not serve as a substitute for explaining one.
Fortress America recognizes another dependency. Hegseth’s stated aim is for military installations eventually to operate under their own power, alongside stronger protection and domestic supply chains. Yet independence from the civilian grid would not by itself eliminate the need for fuel, spare parts, communications, or skilled maintenance. Resilience must extend beyond the perimeter fence.
The same principle applies at the front. Engineering and logistics deserve attention alongside strike footage. They show how forces sustain movement and support combat units, rather than merely how they destroy a target.
From Remote Control to Robotic Forces
The next stage of warfare is unlikely to resemble an army of humanoid machines replacing soldiers. A more plausible future is a force in which people fight alongside growing numbers of specialized robots, from reconnaissance aircraft to ground vehicles carrying ammunition, evacuating casualties, and supporting assaults.
That transition is already underway. Both Russian and Ukrainian forces are using ground robots for logistics, evacuation, and other battlefield tasks. In September, Ukraine’s Defense Ministry reported that its ground systems had completed more than 25,000 supply and evacuation missions in August alone. The figure is an official claim, not an independently audited total, but it illustrates how far the ambition extends beyond occasional demonstrations.
Over the next five years, the important change may be the organization of these machines into working formations. Aerial drones could scout ahead while ground robots move supplies and support troops. Humans would still set objectives and handle situations the machines cannot reliably interpret. The advantage would come from coordinating those different roles, not simply putting more unmanned vehicles on the battlefield.
Remote control, however, is not autonomy. A vehicle that needs continuous steering from an operator remains dependent on communications and human attention. Greater autonomy could allow machines to navigate, avoid obstacles, or complete limited tasks when a connection fails. None of those functions automatically makes a robot capable of independently identifying legitimate targets or deciding when to fire.
The likely result is a race between robotic capability and countermeasures. More machines may reduce the number of soldiers exposed on supply routes, but those machines will themselves become targets. Communications, batteries, maintenance, and replacement parts will remain constraints. Robotization moves some risks away from people; it does not abolish the costs of war.
Nor will inexpensive robots make heavy weapons obsolete. Small drones, armored vehicles, artillery, and aircraft perform different jobs. The stronger force will be the one that combines them effectively and changes that combination faster than its opponent.
Meridian should be judged against this prospect, not the grandeur of its language. Can Washington build systems that cooperate, survive disruption, and remain affordable enough to replace? Can its procurement machinery adapt as quickly as the battlefield demands?
The next war will not be won by whoever unveils the most impressive robot. It may favor the side that can lose machines without losing momentum, preserve human judgment without slowing every decision, and turn battlefield lessons into production before the enemy does. The real revolution is not a machine that looks like a soldier. It is an army that learns to fight differently because machines are now part of its ranks.
MORE ON THE TOPIC:
