■ Path planning technique:
In order to realize the robot zero teaching, the task can be completed in the real environment.
Truman independently developed the robot real-time motion planning system, considering the kinematics limitation of the robot, the real environment (collision detection) and the work to be completed, completely dynamic. Plan the robot motion path so that the robot can work utonomously with the shortest possible and smooth path. ■ Grasping position planning technique: In the robot work task, especially when the work target position is random, the posture of the robot enters the complex work position, which is a problem. Due to the environment and the kinematics of the robot itself, some postures of the robot Will not be available (eg collision, kinematics unreachable, etc.). Truman's R&D posture planning technology can overcome this difficulty and ensure that the robot works in a reasonable posture. ■ Sucker fine control technology: In the random grab or palletizing, unloading application, the suction cup is a common fixture shape, which is different from the usual unified control of all the suction cups. Chumen's intelligent suction cup control technology considers the size and target placement of the caught items. The shape of the gesture, the collision in the process of grabbing and other factors, intelligent control of which suction cups are used to complete the grab work. In order to solve the performance problem of traditional robot inverse kinematics calculation, Truman independently developed the high-speed robot inverse kinematics algorithm program, so that the inverse solution calculation time is almost negligible in real-time planning problems. And Truman's inverse solution operation has its own motion optimization configuration, so that all inverse solutions on the path are continuous and avoid the kinematic error (singular point) region. So far, the algorithm has been successfully applied to all of Truman's solutions. On the basis of the feedback of the visual system, Truman has the technology of
detecting static (stationary, constant) and dynamic (moving, changing)
collisions with high speed and high precision, which can greatly enhance
the static perception of the robot during the working process. The working environment and its ability to change.  ■ Phase-based binocular vision imaging technology Features: Using the object CAD model as a matching template, find the target object by calculating the similarity in the point cloud of the unordered part Features: For non-standard objects (boxes of different sizes, water bottles, etc.), combined with 2D images and 3D point clouds, assist in segmentation of point clouds by identifying surface features, and complete 3D gesture recognition ■ Fast and high precision robot hand eye calibration algorithm |
Browse mobile website