@proceedings {1892, title = {Simply Grasping Simple Shapes: Commanding a Humanoid Hand with a Shape-Based Synergy}, year = {2017}, month = {12/2017}, address = {Puerto Varas, Chile}, abstract = {

Despite rapid advancements in dexterity and mechanical design, the utility of humanoid robots outside of a controlled laboratory setting is limited in part due to the complexity involved in programming robots to grasp common objects. There exists a need for an efficient method to command high degree-of-freedom (DoF) position-controlled dexterous manipulators to grasp a range of objects such that explicit models are not needed for every interaction. The authors propose a method termed geometrical synergies that, similar to the neuroscience concept of postural synergies, aims to decrease the commanded DoF of the humanoid hand. In the geometrical synergy approach, the method relies on grasp design based on intuitive measurements of the object to be grasped, in contrast to postural synergy methods that focus on the principal components of human grasps to determine robot hand joint commands. For this paper, a synergy was designed to grasp cylinder-shaped objects. Using the SynGrasp toolbox, a model of a twelve-DoF hand was created to perform contact analysis around a small set of cylinders dened by a single variable, diameter. Experiments were performed with the robot to validate and update the synergy-based models. Successful manipulation of a large range of cylindrical objects not previously introduced to the robot was demonstrated. This geometric synergy-based grasp planning method can be applied to any position-controlled humanoid hand to decrease the number of commanded DoF based on simple, measureable inputs in order to grasp commonly shaped objects. This method has the potential to vastly expand the library of objects the robot can manipulate.

}, keywords = {Dexterous Hand, Grasp, Humanoid, Manipulation, Synergy}, attachments = {https://mahilab.rice.edu/sites/default/files/publications/farrell2017simply.pdf}, author = {Logan C. Farrell and Troy A. Dennis and Julia A. Badger and Marcia K. O{\textquoteright}Malley} } @proceedings {1826, title = {SOM and LVQ classification of endovascular surgeons using motion-based metrics}, year = {2016}, month = {01/2016}, pages = {227-237}, abstract = {

An increase in the prevalence of endovascular surgery requires a growing number of proficient surgeons. Current endovascular surgeon evaluation techniques are subjective and time-consuming; as a result, there is a demand for an objective and automated evaluation procedure. Leveraging reliable movement metrics and tool-tip data acquisition, we here use neural network techniques such as LVQs and SOMs to identify the mapping between surgeons{\textquoteright} motion data and imposed rating scales. Using LVQs, only 50\ \% testing accuracy was achieved. SOM visualization of this inadequate generalization, however, highlights limitations of the present rating scale and sheds light upon the differences between traditional skill groupings and neural network clusters. In particular, our SOM clustering both exhibits more truthful segmentation and demonstrates which metrics are most indicative of surgeon ability, providing an outline for more rigorous evaluation strategies.

}, issn = {978-3-319-28517-7}, doi = {https://doi.org/10.1007/978-3-319-28518-4_20}, url = {https://link.springer.com/chapter/10.1007/978-3-319-28518-4_20}, attachments = {https://mahilab.rice.edu/sites/default/files/publications/WSOM_2016.pdf}, author = {Kramer, B.D. and Dylan P. Losey and Marcia K. O{\textquoteright}Malley} }