An edible robot from Swiss researchers at the Ecole Polytechnique Fédérale de Lausanne is taking cues from nature with a bio-inspired design, allowing for remote sensors with no environmental concerns.
Non-toxic and biodegradable materials make the tiny boat-shaped robot extremely environmentally friendly. Only 5 cm long, the small robot is composed of a dense compound based on commercial fish food.
A phenomenon called the Marangoni effect helps propel the tiny robot, involving changes in surface tension that allow mass to travel across a medium. In nature, some aquatic insects rely on this method for their locomotion.
In the case of the tiny robot, a small detachable chamber generates carbon dioxide gas, which then travels along a fuel channel filled with propylene glycol, which is expelled as a propellant. In its current iteration, the boatlike robot can move freely for up to several minutes, driven by the chemical reaction generated by combining citric acid and sodium bicarbonate, similar to the compounds used in model volcanoes created for school projects.
“While the development of miniature swimming robots for natural environments has progressed rapidly, these typically rely on plastics, batteries, and other electronics, which pose challenges for mass deployment in sensitive ecosystems,” says lead author and EPFL PhD student Shuhang Zhang. “In this work, we show how those materials can be replaced by completely biodegradable and edible components.”
A Tough But Edible Robot
Dario Floreano in the School of Engineering led the team from the Laboratory of Intelligent Systems in designing the new robot. In their eco-conscious design, the Swiss team went beyond just making the robot safe for aquatic creatures, they designed it to aid watery ecosystems. The boat’s exterior is comprised of a fish food compound, modified to contain 30% more protein and 8% less fat, providing the edible robot with a tough, resilient cover.
While it may be built tough, when the small device comes to the end of its lifespan, sea creatures will be able to consume it, providing a source of nourishment instead of more waste that makes its way into the environment.
Presently, two variants, one designed to make left turns and another designed to make right turns, were prepared through modifications to the fuel channel. The intention is not to precisely direct the robots to a specific point, but to spread the robots across a surface. The semi-random locomotion makes them appear like natural insects to marine life, aiding their utility for feeding or medicating fish. While no animal testing was incorporated into the initial study, the Swiss team speculates these robots could even be used to stimulate aquatic pets’ cognitive development.
Floreano’s engineers foresee large-scale utilization projects for their edible robot. Use cases include many environmental data applications, such as observing water pH, temperature, pollutants, and microorganisms. Future operators could recall the devices to extract data manually or operate them as remote sensors.
Pursuing Environmentally Sound Robotics
Work remains underway in the field of edible robotics, and right now, the Laboratory of Intelligent Systems is behind several developments, including crop growth monitors, edible computer circuits, and food manipulators. Floreano coordinates the RoboFood consortium, a group with € 3.5 million in funding from the EU to pursue edible robot development.
“The replacement of electronic waste with biodegradable materials is the subject of intensive study, but edible materials with targeted nutritional profiles and function have barely been considered, and open up a world of opportunities for human and animal health,” Floreano says.
The paper “Edible Aquatic Robots with Marangoni Propulsion” appeared on May 7, 2025, in Nature Communications.
Ryan Whalen covers science and technology for The Debrief. He holds an MA in History and a Master of Library and Information Science with a certificate in Data Science. He can be contacted at ryan@thedebrief.org, and follow him on Twitter @mdntwvlf.