Czech scientists have developed swarms of microrobots capable of removing up to 94 percent of microplastics from water and 81 percent from soil. A team from the Technical University of Ostrava created these microrobots, which are moved by an external magnetic field and do not contain motors or batteries. These robots capture microplastics in both water and soil. In one hour, they removed 94 percent of polystyrene from a water sample. The tests have so far only taken place in laboratory conditions, and the nickel they carry poses a problem. Filtering water is a known process: it is pushed through a membrane, and anything too large remains trapped. A handful of soil is a different matter. Plastic fragments slip between mineral grains and organic matter, attach themselves to them, and do not leave. No filter can reach them there. This is where the idea of Czech chemists comes in: instead of passing the environment through a device, send the device into the environment. Microrobots without motors, guided by a magnet. These microrobots have no mechanical components. They are grains of MXene, a layered material whose ultra-thin plates stack like the pages of an open book, onto which researchers have attached 500-nanometer nickel nanoparticles. No motor, no battery, no electronics: a rotating magnetic field applied from the outside causes the nickel to rotate and drives each robot into a tumbling motion, at a maximum speed of 22 micrometers per second. The interest in MXene lies in its surface. Its layers carry chemical groups that attract plastic through electrostatic attraction and weak bonds: when a robot hits a particle, it sticks to it. A magnet brought close at the end of the process retrieves the fleet and its cargo in one motion. The work, signed by Jeonghyo Kim, Apabrita Mallick, Su-Jin Song, and Martin Pumera, was published on August 1, 2026, in NPG Asia Materials, a journal of the Nature group. A swarm that stirs water and soil. The number changes everything. Released in thousands, the devices rotate in unison and generate tiny vortices that draw dispersed particles toward the swarm. In front of a fragment too large for one of them, several work together to encircle it and then transport it. In soil, their collective rotation creates shear forces that detach plastics stuck between aggregates, which a stationary material does not do. The numbers, obtained in one hour of experimentation, are clear. In water, the robots remove 94 percent of polystyrene and 89.2 percent of PET, the plastic used in bottles. In a water-saturated model soil, 80.6 percent and 72.2 percent. The most telling comparison is with the same MXene left stationary, which passively captures what comes within its reach: 81.1 percent and 74.4 percent in water, 13 points less on polystyrene. The problem of nickel. Two model plastics, in a test tube, for one hour. Nature, however, mixes polyethylene, textile fibers, tire dust, and paint residues, often covered with biofilm. The authors state that their work has not been tested under real conditions or on repeated use cycles. Nickel is the other concern, and it is serious. As it corrodes, it releases toxic ions for soil microorganisms and plants, and magnetic recovery is never guaranteed at 100 percent. A lost robot means a metal pollution substituted for a plastic pollution. The researchers therefore call for measuring these leaks and seeking less aggressive magnetic materials for the environment before any deployment. In short, these swarms will not go into rivers. Their domain is the closed enclosure: wastewater treatment plant outflows, industrial effluents, excavated soils treated in tanks. The demand, however, should come: the European Union already aimed to reduce by 30 percent the release of microplastics into the environment by 2030, and Regulation 2025/2365 of November 12, 2025, requires industry to prevent the loss of plastic granules.