ARWED is an upper-body immersive virtual reality system developed for human motion training and rehabilitation. We use Oculus VR goggles together with a Leap Motion sensor to include hand and arm motion in the system. The VR environment can track hand position and also reflect the hand motion for retraining through the mirror neuron system.
As a part of ARWED we have studied the upper-body motion using IR, marker-based motion capture. We are creating accurate models of complex human joints and study upper-body motion synergies for subjects with motion impairment. VR training environments are available and are being tested for post-stroke patient recovery. ARWED has also been extended to an immersive suite for the design of robots that follow human-specified trajectories.
See publications resulting from this project.
Synthesis of the upper-limb kinematic chain
The kinematic structure of the human shoulder have always been some of the most difficult parts to model in the human muscoskeletal system. The shoulder complex is a sophisticated region of the human body that provides vast ranges of motions due to its redundant structure while serving a pivotal role in all arm movements. Several kinematic chains have been proposed an analyzed for the shoulder complex. In general, a mechanism that anatomically resembles a shoulder is not necessarily the best solution; another approach may be to find the model based on the shoulder motion. The GH joint is kinematically and anatomically accepted as a spherical pair, whose location is the output of the shoulder mechanism. Thus, an accurate estimation of the GH joint location is crucial for the shoulder kinematics. We have worked to determine the location of the GH joint using a new method based on scapula rhythm. Some results can be seen here:
Center of rotation for the proximal arm.
Link-based optimization to find optimal anchoring
points on axes.
Run CAD software macro.
Points required by the macro are indicated.
SolidWorks macro windows.
The joint axes are drawn in the solid modeling software, with anchoring points on each axis.
Library parts are imported, fitted to joint axes and assembled for the final design.
The draft paper below contains the summary of the method and was used as base for the paper presented at the IDETC 2015 conference.
Draft: A Design Implementation Process for Robotic Hand Synthesis
Implementation: This SolidWorks macro is to be used in conjunction with the library of parts and an excel file containing the points of the axes. An example excel file is provided here .
The link anchoring points can be manually selected or they can be optimized using for instance our link-based optimization technique. The following Mathematica file can be used to create the Matlab functions to optimize those points.
Or you can watch the video tutorial below for information on the automatic process.
More information: Visit PhD candidate Neda Hassanzadeh webpage.