No abstract

Grasp estimation is a fundamental technique crucial for robot manipulation tasks. In this work, we present a scene-oriented grasp estimation scheme taking constraints of the grasp pose imposed by the environment into consideration and training on samples satisfying the constraints. We formulate valid grasps for a parallel-jaw gripper as vectors in a two-dimensional (2D) image and detect them with a fully convolutional network that simultaneously estimates the vectors’ origins and directions. The detected vectors are then converted to 6 degree-of-freedom (6-DOF) grasps with a tailored strategy. As such, the network is able to detect multiple grasp candidates from a cluttered scene in one shot using only an RGB image as input. We evaluate our approach on the GraspNet-1Billion dataset and archived comparable performance as state-of-the-art while being efficient in runtime.

No abstract

Aimed at the problem that experimental verifications are difficult to execute due to lacking effective experimental platforms in the research field of multi-robot formation, we design a simple multi-robot formation platform. This proposed general and low-cost multi-robot formation platform includes the indoor global-positioning system, the multi-robot communication system, and the wheeled mobile robot hardware. For each wheeled mobile robot in our platform, its real-time position information in the centimeter‑level precise is obtained by the Marvelmind Indoor Navigation System and orientation information is obtained by the six-degree-of-freedom gyroscope. The Transmission Control Protocol/Internet Protocol (TCP/IP) wireless communication infrastructure is selected to support the communication among robots and the data collection in the process of experiments. Finally, a set of leader–follower formation experiments are performed by our platform, which include three trajectory tracking experiments of different types and numbers under deterministic environment and a formation-maintaining experiment with external disturbances. The results illustrate that our multi-robot formation platform can be effectively used as a general testbed to evaluate and verify the feasibility and correctness of the theoretical methods in the multi-robot formation. What is more, the proposed simple and general formation platform is beneficial to the development of platforms in the fields of multi-robot coordination, formation control, and search and rescue missions.