The thermodynamic equilibrium assumption often invoked in modeling ion migration in solid-state materials remains insufficient to capture the true migration behavior of Li ions, particularly in less-crystalline superionic conductors that exhibit anomalously high Li ion conductivity. Such materials challenge classical frameworks and necessitate a lattice dynamics-based perspective that explicitly accounts for nonequilibrium phonon interactions and transient structural responses. Here, we uncover a phonon-governed Li ion migration mechanism in garnet-structured superionic conductors by comparing Ta-doped Li_6.6La₃Zr_1.6Ta_0.4O₁₂ (LLZTO4) to its undoped analogue, Li_6.24La₃Zr₂Al_0.24O_11.98 (LLZO). Through a synergistic combination of terahertz time-domain spectroscopy (THz-TDS), ⁷Li magic-angle spinning nuclear magnetic resonance (MAS-NMR), and Raman spectroscopy, we show that Ta doping softens the host lattice and enhances anharmonic phonons, enabling collective, thereby multi-ion migration beyond the limit of single-ion hopping models. This lattice softening induces a dynamically disordered energy landscape that lowers activation barriers and yields Li ion conductivities approaching those of liquid electrolytes. Our findings demonstrate that anharmonic lattice vibrations can serve as the driving force for ultrafast Li ion migration in solid electrolytes. This paradigm shift establishes a fundamental link between lattice thermodynamics and superionic conduction, providing a conceptual and experimental framework for the design of highly conductive solid-state electrolytes.

Mobile robots operating in outdoor environments face the challenge of navigating various terrains with different degrees of difficulty. Therefore, traversability estimation is crucial for safe and efficient robot navigation. Current approaches utilize a robot's driving experience to learn traversability in a self-supervised fashion. However, providing sufficient and diverse experience to the robot is difficult in many practical applications. In this paper, we propose a self-supervised traversability learning method that adapts to challenging terrains with limited prior experience. One key aspect is to enable prioritized learning of scarce yet high-risk terrains by using a risk-sensitive approach. To this end, we train a neural network through a risk-aware instance weighting scheme. Another key aspect is to leverage traversability pseudo-labels on the basis of a self-training scheme. The proposed confidence-regularized self-training generates high-quality pseudo-labels, thereby achieving reliable data augmentation for unexperienced terrains. The effectiveness of the proposed method is verified in extensive real-world experiments, ranging from structured urban environments to complex rugged terrains.