My dissertation research has largely focused on the design, fabrication, and controls of EMBUR (EMerita BUrrowing Robot), the first legged robot capable of self-burrowing in granular materials.  EMBUR's design draws inspiration from the Pacific mole crab Emerita analoga, a burrowing crustacean which utilizes sets of alternately rotating leg pairs to excavate substrate in the swash zones of the California coast. The design of EMBUR integrates novel hybrid soft/rigid mechanisms to enable operation in granular environments. 

I am currently managing a team of three Masters of Engineering students through the redesign of the robot, "EMBUR 2.0."  We aim to increase burrow depth through a combination of improved mechanical design and more intelligent control strategies.

The geometry and control strategies of EMBUR were informed by granular models- in particular, Granular Resistive Force Theory (RFT).  RFT is a reduced-order empirical modeling technique which greatly reduces the computational costs of more robust simulation techniques such as Discrete Element Methods (DEM).  However, until very recently it could only be applied to two dimensional geometries and trajectories. The initial years of my PhD focused on the expansion of this model to three dimensions (3D RFT), and the creation of easily accessible open source code to aid its implementation.

As part of a research project in collaboration with NASA Johnson Space Center, I worked on the fabrication and field deployment of teams of tethered robots. We demonstrate that by pathing small mobile robots around natural objects, we are able to harness the capstan effect and amplify the attachment forces of these small deployable agents. For field experiments, we used the MiniRHex Open Source Platform as small deployable agents (credits to CMU Robomechanics Lab), and a larger rover built in our own lab as the central payload.  We are able to demonstrate the amplification of holding force of a low traction platform in sand of over 700x.

As an undergraduate researcher in the MIT D'Arbeloff Lab, I worked on modeling, control strategies, and later design features of the Supernumerary Robotic Limbs (SRLs). This wearable robotic device gives its user two additional robotic arms, and is intended to support humans in a variety of tasks, including aircraft manufacturing, construction tasks, and gait rehabilitation. My work focused on designing SRL algorithms for sitting and standing assistance for the elderly. In later years, I worked on the design of a braking mechanism for the telescoping actuation of the SRLs. 

As part of the graduate course ME292C at UC Berkeley (Advanced Engineering Design Graphics), I worked with a team to design an off-road fully suspended motorized handcycle similar to the Jeetrike. The handcycle is suitable for most people with a range of disabilities, including those with quadriplegia. Design completed in PTC Creo and graphics and animation (shown below) were done in Autodesk 3D Studio Max.