Most flying machines are defined by their moving parts: propellers, rotors, flapping wings. The robot described in Nature Communications this week has none. Weighing 36.7 milligrams, it generates thrust from ion wind, the stream of charged air molecules produced when a high voltage ionizes the air around an electrode. There are no mechanical actuators, no bearings, no gears, and the propulsion system is silent and solid-state. In tethered tests it hovered for an hour while holding its orientation under closed-loop control, with enough excess thrust to carry sensors.
Ion wind propulsion, also called electrohydrodynamic thrust, works on a simple principle: a sharp electrode at high voltage ionizes nearby air molecules, and those charged molecules accelerate toward a second electrode, dragging neutral air along with them. The resulting airflow is the thrust. The phenomenon has been known for a century and has flown small fixed-wing aircraft, but applying it to an insect-scale robot is a different problem. Two obstacles stood in the way: thrust too weak to lift useful payloads, and control of flight in multiple degrees of freedom that was nearly impossible with such lightweight, flexible structures.
The team, led by Wei Li at Nanjing University of Posts and Telecommunications with collaborators in Nanjing, Hangzhou, Wuhan, and at the University of Vermont, solved both problems. The key is an origami-inspired manufacturing approach that builds the robot from layered metal-polymer composites, folded into a rigid airframe that weighs almost nothing but holds its shape. The design separates the functions of structure, electrode, and sensor into the folded layers, and the whole robot can be assembled rapidly and cheaply, at what the authors describe as disposable cost.
The numbers demonstrate the control breakthrough. With an inertial measurement unit feeding a closed-loop controller, the robot reduced the root mean square error of its pitch angle by 83% and its roll angle by 89% compared with open-loop operation. It can maintain a stable hover, correct its attitude in flight, and carry a payload including a high-fidelity image sensor and a fiber Bragg grating sensor for strain measurement, all while performing pre-programmed tasks. The 5:1 thrust-to-weight ratio, unusually high for a robot this small, is what makes the payload possible.
The implications concern the scale of robotics rather than a single machine. A robot that weighs less than a gram, has no moving parts, and can be manufactured cheaply enough to be disposable is a candidate for deployment in swarms: dozens or hundreds released into a collapsed building, a damaged industrial plant, or a hazardous environment to map the space, identify materials, and relay information. The authors frame the work as an approach to autonomous microrobotic swarm flight, with potential applications in confined-space surveillance, disaster rescue, and hazardous environment exploration.
The physics of ion wind gives the design unusual robustness. With no moving parts, there is nothing to wear out, jam, or break in the conventional sense, and the propulsion works at small scales where conventional rotors become inefficient. The trade-off is the high voltage required: generating ion wind needs kilovolt-scale potentials, and in this demonstration the power and control signals arrive through a tether, which also allowed the hour-long hover. Untethered operation, with onboard power, remains the frontier.
The caveats are those of a proof of principle. The hour-long hover was tethered, the closed-loop demonstrations establish controllability but not full autonomous navigation, and the transition from a single controlled robot to a coordinated swarm requires solving communication, power, and collision-avoidance problems that are not addressed here. The authors acknowledge that the work introduces an approach rather than a finished system.
The ion-wind route sidesteps the actuation problem entirely by eliminating actuation, and the origami-inspired fabrication, with its metal-polymer composites and rapid assembly, addresses the cost and manufacturability barriers that have kept microrobots in the laboratory. The authors frame the manufacturing advance as overcoming a key barrier to practical swarm deployment.
Scale limits the approach’s reach. Ion wind thrust scales favorably at small sizes, which is why the approach works for a 36.7-milligram robot, but the physics becomes less favorable as machines grow, and the technique is unlikely to power a conventional drone. The niche is the very small, the expendable, and the inaccessible, where a silent, solid-state machine with no moving parts has advantages no propeller-driven design can match.
The milestone is real. Controlled flight of a high-thrust, ultralight, ion-propelled robot with integrated sensing is a step toward aerial machines that have more in common with dust motes than with drones: silent, solid-state, cheap enough to lose, and small enough to go where no one would send a helicopter.
Sources: Tao, Q., Gu, Y., Wang, X. et al. Controlled flight of high-thrust ultralight ion-propelled microrobot with integrated sensing. Nature Communications (2026). DOI: 10.1038/s41467-026-76462-y.