In the not-too-distant future, robotics research and teaching will take a significant step forward thanks to some new R&D work currently underway in Quanser's robotics division. The research involves having a prototype of a future Quanser product - a small Unmanned Ground Vehicle (UGV) - being tele-operated through a gesture-sensing glove linked to a magnetic tracking system.
This will open up new and better possibilities for robotic control research and teaching. Researchers, instructors and students can expect a deeper, more intuitive experience as well as a significantly shorter learning curve. Essentially, this new system will extend the capability of our UGVs for research and teaching by adding a new layer of gesture-based functionality.
Click below to view the tele-operation in action. As you will see, the kinematics and Jacobian of the arms are solved. The hand motion and gestures are calculated, mapped in a global frame, and transmitted wirelessly to the UGV rover. The arm is clutched with the operator's thumb and his index finer controls the gripper.
At the macro level, here's how the system and glove "fit" together: a magnetic motion tracking system and the gesture-sensing glove have been integrated into QUARC control software functionalities. This high resolution tracking system computes the translation and rotation motions, i.e., roll, pitch and yaw, in a pre-defined Cartesian frame. The data is used to compute a transformation matrix and conveys sufficient information about the operator's hand motion. The glove itself contains strain gauges that capture the operator's hand gestures.
Using QUARC communication blocks, the transformational matrix and the glove data are transmitted to the Gumstix processor onboard the small UGV. Infrared sensors and an RGB camera are some of the other devices onboard the UGV. The QUARC program receives the motion commands from the station PC.
The kinematics and Jacobian mapping motions of the robotic arm are computed and the commanded motions are translated into joint level PWM inputs for the arm. The PWM commands are applied to the servos using HiQ. (The latter is a data acquisiton board specially designed and manufactured by Quanser to be used onboard unmanned aerial vehicles and small unmanned ground vehicles.)
This project is the result of coordinated research and contributions from the Quanser Robotics Team. Amin Abdossalami, R&D Engineer, was responsible for the controls, kinematics and tele-operations. Cameron Fulford, Engineering Manager, Systems & Control, designed the hardware interface and made it a module inside QUARC. Don Gardner did the final assembly of the robot and shot the video demonstration.
The small UGV with glove tele-operation functionality will join the fleet of Quanser unmanned systems in the near future. We're very excited about its ability to offer researchers and students a better tool with which to work and learn.