Owned mechanical changes aimed to increase the use-life of a Hugo RAS subsystem from 250 surgical procedures to 1000, including root cause analyses, rapid prototyping / iteration, and testing while minimizing software, clinical, and manufacturing impact

Co-led mechanical updates to reduce subsystem cost by 25% by leveraging multi-body Ansys FEA simulations, tolerance analyses given GD&T, engineering testing, and DFA conversations with our contract manufacturer

Led dozens of vendor DFM conversations to implement mfg changes across sheet metal bent, injection molded, die cast, and machined parts

Investigated 12+ failure modes accounting for 170+ field returns since product launch, root causing those making up 50%+ of them

Revamped the design of the surgical navigation system, reducing weight by 40 lbs, increasing stability parameters 4-fold, and decreasing manufacturing cost by 20%

Description

First-authored journal paper proposing novel frame-to-frame tracking technique of ultrasound features requiring less initial manual work, but matching accuracy Deep Neural Network computer vision algorithms, published in the BioMedical Engineering Online journal

Designed and conducted human study with 26 subjects, analyzing physiologic data from 20+ skin-adhered sensors and markers

Developed Python code repository for MIT Immersion Lab to facilitate the post-processing and feature-tracking of ultrasound videos

Used signal processing techniques (full and windowed fast Fourier transforms, Welch’s method) of accelerometer signals for tremor detection

Designed a robust cycle-testing mechanical fixture for Myomo's powered orthotic system using low-cost and 3D-printed components. Modeled, 3D-printed, and redesigned 5 robotic system attachments to improve ease-of-use. Delivered a certified testing protocol covering fatigue, excessive deflection, and impact tests.