Roboticist argues robots should give back to nature, not just take from it

Barbara Mazzolai started her career as a biologist, studying marine biology at the University of Pisa before working on how heavy metals like mercury move through water, soil, and living things at the Italian National Research Council. A chance opening led her into engineering: a university was recruiting biologists to help design environmental monitoring devices, and she took the job in 1999, working under bioroboticist Paolo Dario on sensors and then robots for monitoring air, water, and soil.

She built her first bioinspired robot with colleagues at Sant'Anna in the mid-2000s: a soft robot modeled on an octopus, meant to show that a soft machine could still apply strong force to its surroundings the way the animal does. She went on to complete a PhD in microsystems engineering at Tor Vergata University of Rome in 2011, and became director of the Center for Micro-BioRobotics at the Italian Institute of Technology that year. She is now the institute's associate director for robotics and heads its Bioinspired Soft Robotics Laboratory.

Mazzolai's research turned toward plants, which she says move constantly through a lifelong process called indeterminate growth: they adapt their shape and behavior to their surroundings, repair themselves, sense, and communicate, even though their motion is too slow or too fast for people to notice easily. She was drawn in particular to how efficiently plant roots explore soil. A root reduces friction by growing only at its very tip, while the thicker base stays still, which uses far less energy than a conventional drill that has to push its entire body forward from above.

To copy that principle, her team built a robot with a miniaturized 3D printer at its tip that feeds heated thermoplastic filament to build a snakelike body behind it as it moves, letting the machine push through soil the way a root does. Sensors in the tip let it avoid obstacles and detect nearby nutrients or water. Mazzolai describes bioinspired design as requiring two separate ways of thinking at once, one biologist's and one engineer's, because a robot is never a direct copy of the organism that inspired it — the living version is both hard to replicate and pursuing different goals.

Other labs are pursuing plant-inspired mechanisms as well. Research on plant movement — including hygroscopic motions in pine cone scales, which open and close by absorbing and releasing water without spending any of the plant's own energy, and the searching motion of vine tendrils and stems — has fed into robot arms, artificial tendrils, and self-adaptive devices used in architecture and medicine, according to a review in the journal published by Springer.

A separate line of research argues that bioinspired robots are needed specifically for conservation work, because conventional tools often damage the environments they are meant to protect. A perspective published in Frontiers in Robotics and AI, and also carried by PubMed Central, notes that ecosystem health has declined sharply over the last 50 years, with an estimated loss of nearly 70% of the planet's biodiversity, and warns of a possible sixth mass extinction. Much conservation work today is led by humans using conventional robots such as wheeled vehicles, propeller-driven underwater vehicles, and multi-rotor drones — tools that struggle on uneven terrain, in tight spaces, or underwater, and that can disturb wildlife with noise or damage sediment and habitats.

The authors argue that robots designed to move and sense more like animals — a subset of what's called bioinspired robotics — could operate closer to plants and animals without disturbing them, opening access to places conventional robots and humans struggle to reach, such as cave systems, the ocean floor, and polar regions. Examples in the review include an underwater robot modeled on crabs that walks the seabed collecting trash and sediment samples, and a soft robotic fish equipped with cameras to record aquatic life while swimming in three dimensions. The same review cautions that this approach still needs collaboration and practical development work before it sees wide use in real conservation settings.

Mazzolai's proposal goes further than making robots less disruptive while they operate — she wants their entire life cycle reconsidered, including what happens after they stop working. She points out that technologies like plastics and car batteries were developed with little thought for their environmental afterlife, and argues robotics should not repeat that pattern. Together with collaborators, she laid out this vision, which she calls sustainability robotics, in a manifesto published in Nature Machine Intelligence in July.

The approach rests on three principles: robots should have minimal environmental impact, should be accessible to people across different regions and income levels, and should be "symbiotic," benefiting both humans and the natural world. Concretely, she wants robots designed so that at the end of their useful life they can be reused, recycled, or biodegraded, rather than discarded. Mazzolai says the ingredients for this already exist, and describes it as a vision that resonates with a new generation of roboticists who want to build technology that responds to environmental damage rather than adding to it.