Universal Robotic Gripping

Creative Machines Lab, Cornell (2009-present)
With colleagues from the University of Chicago and iRobot, I designed and built the "Universal Jamming Gripper" that was featured on the cover of the Proceedings of the National Academy of Sciences (PNAS) in November 2010. This gripper exploits the jamming phase transition of granular materials as the gripping mechanism. Consisting of a mass of granular material encased in an elastic membrane, the gripper passively conforms to the target object, then vacuum-hardens to generate a gripping force. This approach proves to be simple, low cost, and well suited for a wide range of gripping tasks. Improvements on the original design, including the added ability to “shoot” gripped objects by fast inflation of the gripper were published in the IEEE Transactions on Robotics (TRO). Current goals include the achievement of more complicated manipulation tasks with jamming grippers. More information can be found at http://creativemachines.cornell.edu/jamming_gripper.

Controlled Granular Jamming

Creative Machines Lab, Cornell (2008-present)
I am currently studying the phase transition jamming phenomenon of granular materials, with a focus on applications for programmable matter. Jamming is a unique property of granular materials that allows them to undergo a fluid-like to solid-like phase transition without a change in temperature. When grains are loosely combined, they yield under shear stress (like a fluid); if they are packed together, they jam in position and resist applied stresses (like a solid). The term programmable matter describes some not-yet-achieved form of matter that would have the ability to transition from its current shape into any desired shape -- reversibly, on command, and under its own power.

Open Source Universal Testing Machine

Creative Machines Lab, Cornell (2011)
Universal testing machines (like those sold by Instron, Tinius Olsen, MTS Systems, and others) are poorly suited for a parallelized testing approach due to their size and cost. As a solution to this problem we designed a low cost, desktop size, open source, universal testing machine that we call freeLoader (shown at left). Each freeLoader can perform two simultaneous material tests, at loads up to 5 kN with +/- 1.8 N accuracy, and at speeds ranging from 2 mm/min to 30 mm/min. Five freeLoader machines can be built for the same cost as one commercial machine -- yielding a tenfold increase in testing throughput. The freeLoader was first published at the 2011 ASME IDETC/CIE Mechanisms and Robotics Conference. More information (including parts lists, build instructions, a user manual, and control software) can be found at creativemachines.cornell.edu/freeloader.

Leg-wheeled Robotic Locomotion

ARM Lab, University at Buffalo (2007-2008)
A leg-wheeled robot locomotes using wheels that are connected to the chassis through a leg-like suspension system. As part of an NSF Research Experience for Undergraduates (REU), I designed and built such a robot (shown at left) based on a specific mechanism that combined four-bar linkages and torsion springs in order to maximize the range of motion of the suspension while minimizing its degrees of freedom. The goal was to develop a robot that showed improved performance in rough or hazardous terrain over robots with more conventional suspension systems.

Autonomous Vehicle Design

UB Robotics, University at Buffalo (2007-2008)
I helped lead the University at Buffalo Robotics Club in the design and construction of an autonomous robot for the 2008 Intelligent Ground Vehicle Competition (IGVC). This competition requires participants to solve open-ended sensing, estimation, control, and locomotion problems, in order to produce a vehicle that can navigate autonomously, subject to the constraints of its operating environment. We placed tenth out of 41 teams at the competition with the vehicle shown here. The vehicle sported a custom-built skid-steer drive train and utilized electric ATV components. It also featured a digital compass, laser range finder, video camera, and differential GPS.

Consumer Robotics Design

DOES Lab, University at Buffalo (2007-2008)
Consumer Robotics is still a field for which the product design and development process is not very mature. I studied consumer robotics development from a product platforming perspective using a wheeled companion robot as a case study. To analyze the potential market I hosted an online survey which was taken by over 200 potential customers. The robot shown here was designed, built, and tested as part of the project.

Multiobjective Visualization

DOES Lab, University at Buffalo (2007-2008)
It is difficult for humans to gain any valuable insight from data that is displayed in more than the usual three spatial dimensions. There are software programs designed specifically for this type of problem, but most were developed for very specific types of data, and few are widely marketed. My research in multiobjective visualization was in the development of a method for reviewing, comparing, and cataloging available software programs. We aimed to connect research communities by revealing novel uses for preexisting software tools in order to promote multidisciplinary breakthroughs. I presented my results at the 2008 AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference.

Green Engineering

DOES Lab, University at Buffalo (2007)
I received a Zimmer Research Scholar Award to study green engineering and the adoption of green practices in the mechanical engineering design process. My research focused on motivations and strategies for implementing environmentally conscious design and manufacturing practices. I was awarded travel grants to present my research at both the 2007 ASME IDETC/CIE Conference and the 2008 NSF CMMI Engineering Research and Innovation Conference.