Remotely powered microrobots are proposed as next‐generation vehicles for drug delivery. However, most microrobots swim with linear trajectories and lack the capacity to robustly adhere to soft tissues. This limits their ability to navigate complex biological environments and sustainably release drugs at target sites. In this work, bubble‐based microrobots with complex geometries are shown to efficiently swim with non‐linear trajectories in a mouse bladder, robustly pin to the epithelium, and slowly release therapeutic drugs. The asymmetric fins on the exterior bodies of the microrobots induce a rapid rotational component to their swimming motions of up to ≈150 body lengths per second. Due to their fast speeds and sharp fins, the microrobots can mechanically pin themselves to the bladder epithelium and endure shear stresses commensurate with urination. Dexamethasone, a small molecule drug used for inflammatory diseases, is encapsulated within the polymeric bodies of the microrobots. The sustained release of the drug is shown to temper inflammation in a manner that surpasses the performance of free drug controls. This system provides a potential strategy to use microrobots to efficiently navigate large volumes, pin at soft tissue boundaries, and release drugs over several days for a range of diseases.

Detection of biomolecules is essential for patient diagnosis, disease management, and numerous other applications. Recently, nano- and microparticle-based detection has been explored for improving traditional assays by reducing required sample volumes and assay times as well as enhancing tunability. Among these approaches, active particle-based assays that couple particle motion to biomolecule concentration expand assay accessibility through simplified signal outputs. However, most of these approaches require secondary labeling, which complicates workflows and introduces additional points of error. Here, we show a proof-of-concept for a label-free, motion-based biomolecule detection system using electrokinetic active particles. We prepare induced-charge electrophoretic microsensors (ICEMs) for the capture of two model biomolecules, streptavidin and ovalbumin, and show that the specific capture of the biomolecules leads to direct signal transduction through ICEM speed suppression at concentrations as low as 0.1 nM. This work lays the foundation for a new paradigm of rapid, simple, and label-free biomolecule detection using active particles.