Real-time localization is a requirement for the clinical application of magnetically actuated wireless capsule endoscopy. This study proposes a physics-informed deep learning framework to address the model-plant mismatch inherent in analytical methods and the data requirements of end-to-end learning. A dual-branch residual network, physics-informed residual network, computes corrections for the magnetic dipole model. Two operational paradigms are evaluated: a lookup-table-based approach and an end-to-end optimization approach. Experiments indicate that the framework achieves a mean position error of 0.69 $\pm$ 0.29 mm and a mean orientation error of 0.70 $\pm$ 0.37$^{\circ }$, utilizing less than 20% of the calibration configurations required by data-driven baselines. Validation includes an ex vivo signal propagation study through biological tissue and compatibility tests with an anti-interference filter, supplemented by a preliminary in vivo animal experiment. The results provide a methodology for sim-to-real transfer in robotic capsule tracking systems.
This study presents a magnetically actuated soft robot inspired by the peristaltic locomotion of flatworms, designed to replicate the biological locomotion of worms to achieve robust maneuverability. Fabricated entirely from photocurable soft resin, the robot features a flexible elastomeric body and two webbed fins with embedded soft magnets. By applying a vertically oscillating magnetic field, the robot achieves forward crawling through the coordinated bending and lifting of fins, converting oscillating magnetic fields into continuous undulatory motion that mimics the gait of flatworms. The experimental results demonstrate that the system maintains consistent bidirectional velocities in the range of 4–7 mm/s on flat surfaces. Beyond linear locomotion, the robot demonstrates effective terrain adaptability, navigating complex topographies, including curved obstacles up to 16 times its body thickness, by autonomously adopting a high-lifting kinematic strategy to overcome gravitational resistance. Furthermore, load-carrying tests reveal that the robot can transport a 6 g payload without velocity degradation. These findings underscore the robot’s efficacy in overcoming mobility constraints, highlighting promising applications in fields requiring non-invasive intervention, such as biomedical capsule endoscopy and industrial pipeline inspection.