How Robot Stingrays Could One Day Sink a Battleship

The Real Robot Stingray That Started the Conversation

In 2016, researchers led by scientists associated with Harvard University published a remarkable experiment in Science. They created a miniature tissue-engineered ray using a flexible silicone body, a gold skeleton, and approximately 200,000 rat heart-muscle cells.

The cells were genetically modified to contract when exposed to pulses of blue light. When the muscle layer contracted, it pulled the robot’s fins downward. The elastic gold framework then helped return the fins to their original position. Repeating that cycle produced the graceful wave-like motion used by real rays.

A Robot Guided by Light

The researchers could steer the device by changing the timing and direction of the light pulses. In demonstrations, the tiny robotic ray followed a course and maneuvered around obstacles. It was an impressive example of optogenetics, tissue engineering, and bio-inspired robotics working together instead of arguing over whose laboratory had the better coffee machine.

The research was primarily intended to improve scientists’ understanding of muscular pumps, including the human heart. It was not a weapons program. The ray’s living cells needed nutrients, controlled conditions, and light-based stimulation. It had no onboard computer, battery, sonar, navigation system, or payload.

Calling this little swimmer a future ship-killer therefore requires an Olympic-class leap. The important point is not that the original machine was dangerous. It is that researchers proved a ray’s propulsion could be reconstructed in a soft, controllable platform.

Do Not Confuse the Raybot With DARPA’s Manta Ray

A separate project makes the naval connection much more concrete. The Defense Advanced Research Projects Agency developed the Manta Ray program to explore long-range, long-duration, payload-capable uncrewed underwater vehicles.

Northrop Grumman’s full-scale Manta Ray prototype completed major in-water testing off Southern California in 2024. According to DARPA, the tests covered submerged movement using buoyancy, propellers, and control surfaces. The modular vehicle was transported in sections, assembled near the test area, and designed with payload bays for different mission equipment.

DARPA’s machine is not a scaled-up version of the heart-cell ray. It does not flap living fins, and its resemblance to a manta is partly a result of hydrodynamic engineering. Still, the two projects sit on the same technological family tree: engineers study nature, identify efficient features, and translate those features into machines.

Why Build an Underwater Robot Like a Stingray?

Efficient, Low-Disturbance Movement

Rays move by passing waves along broad pectoral fins. This method can provide excellent control without relying exclusively on rapidly spinning propellers. A robotic version could potentially create less turbulence and a different acoustic signature than a conventional torpedo-shaped vehicle.

MIT’s soft robotic fish, SoFi, demonstrated the value of lifelike underwater propulsion for peaceful research. The untethered robot used an undulating tail, acoustic controls, a camera, and three-dimensional swimming to observe marine life at depths of up to 18 meters. Its movements allowed it to approach fish without causing the disruption associated with some traditional underwater machines.

A military robot would have a different purpose, but the physical advantage carries over. A vehicle that disturbs less water and produces less familiar machinery noise may be harder for passive sensors to recognize.

Maneuverability Near the Seafloor

A flat robot with flexible fins could be useful around reefs, harbor structures, wreckage, cables, and uneven seabeds. These are awkward environments for large submarines and fast-moving torpedoes. They are also strategically important places where mines, sensors, communications lines, and naval infrastructure tend to live.

Soft structures could help a robot tolerate minor contact and squeeze past obstacles. A low profile might also allow it to rest on the bottom while conserving energy. For long missions, patiently waiting can be more valuable than racing around like an underwater sports car with something to prove.

A Less Obvious Sonar Target

Sonar does not simply display a helpful label reading “EVIL ROBOT, 300 YARDS.” Operators interpret reflected sound, motion, size, and behavior. A small, slow, soft-bodied vehicle may present a less familiar signature than a metal cylinder moving directly toward a ship.

Biomimicry would not make an underwater drone invisible. Active sonar, magnetic sensors, optical systems, trained marine mammals, defensive drones, and improved classification software could still reveal it. However, forcing defenders to separate robots from animals, debris, and background noise would increase their workload.

How Robot Stingrays Could Threaten a Warship

The most realistic danger is not one giant mechanical stingray delivering a cinematic knockout blow. It is a distributed network of affordable underwater drones contributing to a larger operation.

Persistent Surveillance

A robotic ray could carry passive acoustic sensors, cameras, environmental instruments, or magnetic detectors. Instead of attacking anything, it might quietly observe a harbor, shipping lane, or anchorage. Information gathered by several robots could help identify vessel movements and changes in defensive patterns.

This scouting role matters because a weapon can only engage a target that has first been located and tracked. Small underwater drones might become roaming components of a wider sensor network rather than miniature torpedoes.

Coordinated Swarms

Harvard’s Blueswarm research has already shown that small fish-inspired robots can coordinate underwater. Each Bluebot used cameras and LED signals to estimate the position of nearby robots. The group demonstrated aggregation, dispersion, circular movement, and a basic cooperative search behavior without a central controller directing every turn.

Future military systems could apply more advanced versions of the same principle. A swarm could spread out to search a larger area, exchange local observations, and reorganize when members fail. Destroying one drone would not necessarily stop the mission because the intelligence and decision-making would be distributed across the group.

Decoys, Distraction, and Defensive Exhaustion

Robot stingrays would not need to carry explosives to cause trouble. Some could generate misleading acoustic or magnetic signals. Others might encourage a ship to activate sonar, launch defensive equipment, change course, or reveal its position.

The cost imbalance is especially uncomfortable for navies. A sophisticated warship may have expensive defensive systems and a limited inventory of countermeasures. If it must respond to many cheap contacts, even harmless decoys can consume attention, time, and resources.

From Mission Kill to Physical Destruction

Military planners distinguish between sinking a ship and achieving a “mission kill.” A vessel does not have to disappear beneath the waves to become operationally useless. Damage to propulsion, sensors, communications, or other critical systems could force it to withdraw for repairs.

An underwater drone could theoretically contribute to a physical attack by identifying a target, placing a sensor, guiding another system, or carrying a military payload. A coordinated group could approach from several directions and complicate defensive tracking. The exact weaponization methods are speculative, classified where they exist, and beyond the harmless laboratory designs that inspired the robot-stingray idea.

Could One Actually Sink a Battleship?

Technically, the answer is “perhaps one day,” followed by a warehouse full of qualifications.

Traditional battleships were enormous armored vessels built to trade heavy gunfire. The United States no longer operates active battleships; the surviving Iowa-class ships are museums. In modern conversation, “battleship” is often used loosely to mean a destroyer, cruiser, carrier, or other major surface combatant.

Modern warships are formidable, but they are not invulnerable. Recent maritime conflicts have demonstrated that comparatively inexpensive uncrewed craft can damage or constrain much more valuable vessels. Ukraine’s use of explosive surface drones against Russia’s Black Sea Fleet accelerated global interest in distributed naval robotics and highlighted the financial advantage of risking machines instead of crews.

An animal-inspired underwater drone might add stealth and persistence to that equation. Yet sinking a large, alert warship at sea would remain extraordinarily difficult. Naval vessels operate with escorts, sonar, radar, helicopters, electronic-warfare equipment, weapons, damage-control teams, and increasingly sophisticated counter-drone systems.

The more plausible near-term outcome is disruption: forcing ships to slow down, leave an area, deploy countermeasures, or spend heavily on underwater security. A robot that never fires a weapon might still alter naval strategy merely by making commanders worry about what is resting on the seabed.

The Engineering Problems Are Still Enormous

Power and Endurance

Underwater vehicles must carry energy or harvest it from their environment. Batteries add weight, while high speed rapidly consumes power. DARPA’s Manta Ray work has explored buoyancy-driven gliding, energy management, corrosion resistance, and long-duration operation precisely because endurance is one of the field’s hardest problems.

Communication Below the Surface

Ordinary radio signals do not travel efficiently through seawater. Underwater robots commonly use acoustic communication, surface periodically, or complete missions with limited contact. Acoustic links have lower bandwidth and greater delays than the connections used by aerial drones. Consequently, underwater autonomy is not a luxury; it is often a necessity.

Navigation Without GPS

GPS signals are generally unavailable while a vehicle is submerged. An autonomous robot may instead combine inertial navigation, depth measurements, sonar, environmental maps, and occasional position updates. Small errors accumulate over time, which becomes a serious problem when a mission lasts days or months.

Biofouling, Salt, and Pressure

The ocean is a rude workplace. Salt corrodes components, pressure tests every seal, currents push vehicles off course, and marine organisms happily colonize exposed surfaces. Flexible fins must survive thousands or millions of motion cycles without tearing. A robot designed to resemble a stingray must still endure conditions that would make an ordinary consumer gadget file for immediate retirement.

Scaling Living Muscle

Biohybrid propulsion creates additional difficulties. Living cells need nutrients, oxygen, waste removal, temperature control, and protection from contamination. A tiny tissue-engineered ray in a laboratory is fundamentally different from an ocean-going vehicle carrying sensors and power systems.

Future biohybrid robots may incorporate engineered muscle, onboard circulation, or synthetic tissues, but most practical military ray designs are more likely to use electric, hydraulic, magnetic, or other artificial actuators. Nature can supply the blueprint without supplying the actual meat.

Autonomy Creates Legal and Ethical Questions

An underwater robot that independently searches for equipment is one thing. A system allowed to select and attack a ship is another. Murky water, incomplete sensor data, civilian traffic, marine animals, and unreliable communications create obvious risks of misidentification.

Meaningful human control, clear rules of engagement, reliable abort mechanisms, cybersecurity, and compliance with the law of armed conflict would all be central to responsible deployment. Swarms make accountability even harder because their behavior emerges from interactions among many machines.

There is also an environmental problem. Lost underwater drones can become electronic debris, leak battery chemicals, interfere with wildlife, or remain active long after a conflict. Building a robot that looks like nature does not automatically make it friendly to nature.

What the First Military Robot Rays Would Probably Do

The earliest operational systems are more likely to conduct surveillance, seabed mapping, infrastructure inspection, mine countermeasures, and anti-submarine sensing than direct attacks. These missions offer valuable information while placing fewer technical and legal demands on autonomy.

Navies may also combine different platforms. A large, long-endurance vehicle could transport or recharge smaller ray-shaped drones. Surface vessels might relay communications, while fish-like robots investigate complex terrain. Aerial drones could provide wide-area observations as underwater units examine contacts below the surface.

This mixed ecosystem is more believable than a single miraculous machine. The future ocean may contain large gliders, small soft robots, fixed sensors, uncrewed boats, and crewed ships cooperating as one network. The “robot stingray” would be a specialist in that network, not Aquaman’s mechanical sidekick.

Experience-Based Perspective: What Working With Robot Rays Would Really Be Like

Accounts from underwater robotics research reveal a recurring lesson: the dramatic demonstration is usually the shortest part of the job. Before a robot glides elegantly past a camera, engineers spend long hours checking seals, balancing buoyancy, calibrating sensors, and chasing software errors that appeared only after the machine got wet.

Consider the experience of preparing a biomimetic robot for an open-water trial. In a laboratory tank, lighting is predictable, water is calm, and the walls provide clear visual references. At sea, sunlight flickers, suspended particles confuse cameras, currents bend the intended route, and acoustic signals bounce from the seabed. A swimming motion that looked flawless indoors may become inefficient when waves and turbulence join the meeting without an invitation.

Operators also learn that neutral buoyancy is unforgiving. A slight imbalance can make a robot gradually sink, rise, or roll. Adding a new camera or battery changes its center of mass. Flexible fins may produce different thrust as materials age or water temperature changes. Teams therefore test one modification at a time and keep recovery equipment nearby. The ocean has an impressive collection of prototypes already, and it does not issue receipts.

The Harvard raybot experience highlights another reality: biological machines require caretaking. Living muscle cannot simply be switched off, stored in a toolbox, and rediscovered six months later beneath a pile of cables. Researchers must maintain suitable chemical and environmental conditions. Even when the cells perform well, engineers must coordinate biology with structural mechanics and external control.

MIT’s SoFi tests illustrate the rewards of solving those problems. A soft robotic fish was able to maneuver near real marine life while recording underwater scenes. For a researcher, that represents more than a charming video. It demonstrates that a robot can gather observations without dominating the behavior of the animals being studied. A quieter, less intrusive platform may reveal activity that conventional vehicles accidentally suppress.

Swarm experiments add a different operational experience. Managing many simple robots can be more frustratingand more powerfulthan operating one sophisticated vehicle. Individual units drift, lose sight of neighbors, or interpret signals differently. Yet when local coordination works, the group can cover territory and recover from failures without waiting for a central commander.

A naval operator would face the same trade-off on a far larger scale. A robot-ray swarm could provide many viewpoints, but it would also produce an avalanche of uncertain contacts. Command software would need to summarize information instead of displaying every sensor reading. Otherwise, the machines would reduce physical workload while creating a new profession: exhausted human staring at 200 blinking icons.

Real maritime conflicts offer the final experience-based warning. Uncrewed systems can create outsized effects, but opponents adapt quickly. Defenders deploy barriers, patrol aircraft, electronic countermeasures, decoys, guns, sonar, and their own drones. A successful design does not remain mysterious forever. Once recovered or observed, it becomes a problem for engineers on the other side to solve.

That is why the robot stingray’s most important advantage may not be its shape. Its value could come from rapid manufacturing, replaceable payloads, shared software, and the ability to learn from repeated deployments. The ray form provides efficient motion and a useful degree of ambiguity, but adaptability determines whether the system remains relevant after the first surprise wears off.

Conclusion

Robot stingrays are not currently prowling the oceans in search of battleships. The famous biohybrid ray was a tiny medical-engineering experiment, while DARPA’s much larger Manta Ray is a separate long-endurance underwater vehicle. Neither provides evidence of an operational animal-like ship-killing weapon.

However, the technologies surrounding them are advancing: soft propulsion, autonomous navigation, underwater communication, cooperative swarms, compact sensors, and long-duration energy management. Combined, those capabilities could produce quiet robotic scouts, decoys, mobile sensor networks, or military underwater drones that challenge traditional naval defenses.

The future threat is therefore not a single mechanical monster rising from the deep. It is a patient, distributed ecosystem of machines that can watch, wait, cooperate, and force expensive ships to defend every directionincluding the apparently empty water beneath them.