At Density, we are seeking to inform workplace (and subsequently, city) design with space analytics data. As a data scientist, I design the algorithms to convert raw sensor data into business value for our customers. This includes using computer vision and point cloud processing techniques to detect people in space from 3D data (point cloud & depth images from millimeter wave radar & time-of-flight sensors), tracking those people through time, then finally showing customers how their workplaces are used with that data.

At Delsys, I led the technical development of pose estimation and body landmark tracking projects using computer vision and 3D geometrical modelling techniques on point cloud & image data. I worked with and learned from some fantastic critical thinkers as we sought to extend the accessibility of precise movement analysis for patients with various neuromusculoskeletal disorders, with the ultimate goal of improving their rehabilitation through these measurements. I presented our research for this project at the International Society of Biomechanics 2020. Delsys is a place with significant opportunity for those ambitious enough to take advantage, and as such I was also able to work on the following projects: - Virtual IMU Realignment Algorithm: Designed an algorithm to virtually realign Inertial Measurement Units (IMUs) on a person's body to optimize the data quality from the sensors- Kinematics Toolbox: Developed a toolbox to process, synchronize, and analyze data from multiple body-worn sensor types (point cloud, image, marker, IMU, etc.) for kinematic analysis- Signal Processing Toolbox: Contributed to the development of a toolbox for real-time signal processing & feature extraction

My thesis focused on designing machine learning algorithms to estimate forces in a person’s knee with a shoe insole. Specifically, I sought to monitor musculoskeletal-modelled knee contact forces from a wearable instrumented insole, with the ultimate goal of improving rehabilitation from ACL surgery and total knee replacement. My thesis included the following projects:- Activity Classification: Developed a machine learning model to predict what activity a person was performing from wearable sensor data using features from time, frequency, and wavelet domains. Using only force-sensing insoles, accuracy was >97%. I presented this work at the International Society of Biomechanics in 2019 and was a finalist for the Young Investigator Award. - Knee Contact Force Prediction: Developed a machine learning model to predict internal knee contact forces from wearable sensor data, which were estimated using musculoskeletal modelling. Using a single force-sensing insole, model accuracy was less than 1/10 of a person's bodyweight. While in graduate school, I pursued 2 Master's Degrees simultaneously, and as a result was also able to complete the following projects:- Post-Surgery Knee Mechanics Device: Built a device & software to assess joint properties of patients' knees following total knee replacement & ACL reconstruction. - ACL Muscle Force Steadiness Algorithm - Developed an algorithm to further understand quadriceps & hamstring adaptation following ACL reconstruction. Published in Clinical Biomechanics, and given research merit-based award at the American Society of Biomechanics in 2018.

At Leoni, my work focused on redesigning wiring harnesses to maximize profit per unit while simultaneously improving performance. My primary responsibilities included working with customers to understand core requirements, optimizing harness design to maximize those core requirements, and presenting at design reviews. I was also able to work internationally during this experience, where I spent 3 months in our manufacturing facility to better understand design concepts that could reduce manufacturing time and costs.