Latest Projects
Dual Arm Robotic Manipulation
This demo is part of the project of "Machine-learning of vision-guided bi-manual grasps, for adaptable autonomous manipulation in manufacturing environments", Sponsored by Cranfield-Loughborough EPSRC. This video describes the implementation, demonstration and evaluation of a variety of autonomous, vision-guided manipulation capabilities, using a dual-arm Baxter robot. Initially, symmetric coordinated bimanual manipulation was implemented on the robot, to achieve task space adaptive control, using quaternion parameters. This coordinated control method enables one arm (master arm) to be moved along a desired trajectory, while the other arm (slave arm) automatically follows the master arm while maintaining a constant relative pose between the two end-effectors.
Dynamic Grasp and Trajectory Planning for Moving Objects
This paper shows how a robot arm and hand can follow and grasp moving objects tracked by a vision system, as is needed when a human hands an object to the robot during collaborative working. While the object is being moved by the human co-worker, a set of likely grasps, generated by a learned grasp planner, is continuously searched to determine which grasp is optimal with respect to both: the current configuration of the robot with respect to the target grasp; and the constraints of finding a collision free trajectory to reach that grasp configuration, while avoiding collisions with the human (also tracked by vision) and other obstacles.
one-shot learning approach for grasping unknown objects
This demo is part of RoMaNS project. A model of each grasp type is learned from a single kinesthetic demonstration and several types are taught. These models are used to select and generate grasps for unfamiliar objects. Both the learning and generation stages use an incomplete point cloud from a depth camera, so no prior model of an object shape is used. The learned model is a product of experts, in which experts are of two types. The first type is a contact model and is a density over the pose of a single hand link relative to the local object surface. The second type is the hand-configuration model and is a density over the whole-hand configuration.
Ankle rehabilitation by a parallel robot
In this project, I investigated the performance of a 6-DoF parallel robot in tracking the movement of the foot trajectory of a paretic leg during a single stride.The foot trajectories of nine patients with a paretic leg including both males and females have been measured and analysed by a Vicon system in a gait laboratory. Based on kinematic and dynamic analysis of a 6-DoF UPS parallel robot, an algorithm was developed inMATLAB to calculate the length of the actuators and their required forces during all trajectories.The workspace and singularity points of the robot were then investigated in nine different cases. A 6-DoF UPS parallel robot prototype with high repeatability was designed and built in order to simulate a single stride.
In-lab testing of autonomous laser-cutting, with various different object shapes, and dummy laser-pointer, commercialised via ERL’s spin-out company A.R.M Robotics Ltd
-
The robot moves the camera to multiple viewpoints, and reconstructs 3D model of the scene.
-
The AI automatically segments the foreground object from background points.
-
The human operator selects a cutting plane using two mouse-clicks.
-
The AI automatically selects the points along the surface to be cut, and plans a trajectory for the laser, including desired laser stand-off distance and automatically maintaining laser axis orthogonal to the local surface curvature throughout.
-
The AI automatically plans a collision free trajectory of the robot’s joints to deliver the laser along the cutting trajectory.
Robofish
In order to control and analyze robotic fish movement, dynamic model and lateral movements of the robotic tail were studied. The posterior tail creates thrust force. The mean rate of work of the lateral movements is equal to the sum of the mean rate of work available for producing the mean thrust and the rate of shedding of kinetic energy of lateral fluid motions. The mean thrust is calculated entirely from the displacement and swimming speed at the trailing edge of the caudal fin.
Dual Arm Robotic Manipulation
This demo is part of the project of "Machine-learning of vision-guided bi-manual grasps, for adaptable autonomous manipulation in manufacturing environments", Sponsored by Cranfield-Loughborough EPSRC. This video describes the implementation, demonstration and evaluation of a variety of autonomous, vision-guided manipulation capabilities, using a dual-arm Baxter robot. Initially, symmetric coordinated bimanual manipulation was implemented on the robot, to achieve task space adaptive control, using quaternion parameters. This coordinated control method enables one arm (master arm) to be moved along a desired trajectory, while the other arm (slave arm) automatically follows the master arm while maintaining a constant relative pose between the two end-effectors.
