Matryoshka doll-like robot changes its shape in real time and in situ

Until now, when scientists created magnetic robots, their magnetisation profiles were generally fixed, enabling only a specific type of shape programming capability using applied external magnetic fields. Researchers at the Max Planck Institute for Intelligent Systems (MPI-IS) have now proposed a new magnetisation reprogramming method that can drastically expand the complexity and diversity of the shape-programming capabilities of such robots. They built a soft robot with a magnetisation profile that can be altered in real time and in situ. Their findings were published in Nature on 11 September 2025.

Led by Prof. Dr. Metin Sitti in the Physical Intelligence (PI) Department at MPI-IS in collaboration with Koç University in Istanbul, Turkey, the team stacked several tubes inside each other like Matryoshka dolls. As can be seen in Figure 1A, tube C is nested inside tube B, which in turn is nested inside tube A.

Each tube contains one or more magnetic units, and the magnetisation profile of each magnetic unit can be pre-programmed on demand (Figure 1B). When the tubes' stacking configuration changes via another non-magnetic actuation method, such as sliding the tubes apart or closer together, the relative position of the magnetic units, and consequently the magnetisation profile of the entire stack, alters (Figure 1C).

This real-time, in-situ generation and transformation of shapes has not been possible with previous magnetic robots. Now however, with the magnetic field kept constant, a tube can change from a straight line to a helix (Figure 1D), or deform in the opposite direction (Figure 1E). Moreover, this approach can be extended to two- and three-dimensional frameworks, enabling real-time switching between multiple deformation modes without altering the magnetic field (Figures 1F and 1G).

While the focus at Max Planck Institutes is primarily on curiosity-driven basic research, the team has also explored how this method could be applied in various scenarios, such as navigating around objects without undesired contact, reprogramming cilia arrays and coordinating multiple instruments either cooperatively or independently under the same magnetic field.

However, this research could also have practical applications one day. For example, in medicine – particularly in minimally invasive, image-guided treatments for vascular diseases. During these procedures, physicians guide a catheter and guidewire through the blood vessels to the target lesion for diagnosis or therapy. As the catheter navigates curved vessels, friction and contact with the vessel wall are inevitable, which can cause damage that delays recovery and, in severe cases, results in medical complications. Older patients in particular often decide against such procedures, opting for medication instead.

The new technology, which has now been published in Nature, offers a compelling alternative: by adjusting the catheter’s magnetisation profile in real time to match the path ahead, friction and contact could be greatly reduced – or even eliminated entirely – when navigating curved vessels. This would minimise damage to delicate tissue, promote faster recovery and make vascular interventions a viable option for patients who would otherwise be excluded from these procedures due to age or vessel fragility.

“This stack of tubes could become the guiding principle of a new catheter technology in the future. While this is basic research at its best, we see high potential for translating this work into diverse real-world applications in the near future,” says Sitti, formerly the Director of the PI Department at MPI-IS and now President of Koç University in Istanbul.

“Our initial goal was to develop a method that could alter a magnetisation profile in real time and in situ,” says Xianqiang Bao, the first author of the publication. “During the research, we discovered unexpected capabilities, such as shape retention and magnetic neutralisation, which open up new possibilities for technologies like catheter design and cilia array reprogramming.”

“This fundamental work offers many potential application scenarios. In our future research, we aim to integrate this method into specific applications and explore its feasibility in other fields,” say Fan Wang and Jianhua Zhang, the two other co-first authors of the publication.

Reference:
Bao, X., Wang, F., Zhang, J. et al. Real-time in situ magnetization reprogramming for soft robotics. Nature 645, 375–384 (2025).

https://doi.org/10.1038/s41586-025-09459-0