ABSTRACT Icing poses a significant challenge in energy, transportation, and (soft) robotic applications, leading to disfunctions and efficiency losses. Traditional de‐icing methods are often energy‐intensive, non‐selective, or difficult to integrate with robotic and autonomous systems. Here, we introduce a 4D‐printing strategy that enables programmable spatio‐temporal de‐icing, providing an energy‐efficient pathway to precisely control when and where ice is removed at the millimeter scale. By integrating raspberry‐like gold–magnetite nanofillers into a photosensitive resin and employing a magnetic‐assisted Digital Light Processing (DLP), we align the fillers into oriented nanoparticle chains, producing anisotropic architectures with dual functionality: magnetic‐driven mobility actuation and plasmonic heating for localized thermal release. Hyperspectral analysis shows that the photothermal response can be finely tuned through plasmonic grafting density, while geometric programming governs the amplitude and spatial distribution of heating. As a demonstration of this multifunctional platform, we develop an autonomous icebreaking microrobot capable of navigating frozen landscapes while melting ice with pinpoint accuracy. This work introduces a new class of 4D‐printed magneto‐plasmonic materials enabling on‐demand phase‐transition control, opening opportunities for energy‐efficient de‐icing, adaptive soft robotics for extreme environments and time‐programmable devices.
Pérot et al. (2026) studied this question.