Active colloids show non-equilibrium behavior that departs from classical Brownian motion, thus providing a platform for novel fundamental phenomena and for enticing possible applications ranging from water treatment to medicine and microrobotics. Although the physics, motion mechanisms and guidance have been extensively investigated, active colloids are rarely exploited to simultaneously guide and transport micron-sized objects in a controllable and reversible manner. Here, we use autonomous active Janus particles as colloidal shuttles to controllably transport cargo at the microscale using external electric and magnetic fields. The active motion arises from the metallodielectric characteristics of the Janus particles, which allows them to also trap, transport and release cargo particles through dielectrophoretic interactions induced by an AC electric field. The ferromagnetic nature of the nickel layer that forms the metallic hemisphere of the Janus colloids provides an additional mechanism to direct the motion of the shuttle using an external magnetic field. With this highly programmable colloidal system, we are able to harness active colloid motion and use it to transport cargo particles to specific destinations through a pre-defined route. A simple analytical model is derived to successfully describe the motion of the shuttle-cargo assembly in response to the applied electrical field. The high level of control on cargo pick-up, transport and release leads to a powerful delivery tool, which could eventually be used in microactuators, microfluidics or for controlled delivery within organ-on-a-chip devices.

Regularized methods have been widely applied to system identification problems without known model structures. This paper presents an infinite-dimensional sparse learning algorithm based on atomic norm regularization. Atomic norm regularization decomposes the transfer function into firstorder atomic models and solves a group lasso problem that selects a sparse set of poles and identifies the corresponding coefficients. The difficulty in solving the problem lies in the fact that there are an infinite number of possible atomic models. This work proposes a greedy algorithm that generates new candidate atomic models maximizing the violation of the optimality conditions of the existing problem. This algorithm is able to solve the infinite-dimensional group lasso problem with high precision. The algorithm is further extended to reduce the bias and reject false positives in pole location estimation by iteratively reweighted adaptive group lasso and complementary pairs stability selection respectively. Numerical results demonstrate that the proposed algorithm performs better than benchmark parameterized and regularized methods in terms of both impulse response fitting and pole location estimation.

Regularized methods have been widely applied to system identification problems without known model structures. This paper proposes an infinite-dimensional sparse learning algorithm based on atomic norm regularization. Atomic norm regularization decomposes the transfer function into firstorder atomic models and solves a group lasso problem that selects a sparse set of poles and identifies the corresponding coefficients. The difficulty in solving the problem lies in the fact that there are an infinite number of possible atomic models. This work proposes a greedy algorithm that generates new candidate atomic models maximizing the violation of the optimality condition of the existing problem. This algorithm is able to solve the infinite-dimensional group lasso problem with high precision. The algorithm is further extended to reduce the bias and reject false positives in pole location estimation by iteratively reweighted adaptive group lasso and complementary pairs stability selection respectively. Numerical results demonstrate that the proposed algorithm performs better than benchmark parameterized and regularized methods in terms of both impulse response fitting and pole location estimation.