Jordan Charest Email & Phone Number

Waltham, Ma, Us

Compute Research Team• Research and experiment with the latest robotics algorithms and hardware, in order to accelerate existing applications, enable new abilities, reduce costs, and improve performance• Ported the Spot codebase to arm64• Redesigned and rewrote Spot mapping software to utilize CPU intrinsics and the GPU. Result: 2.5x increase in observable area, 30% faster processing time, 50% decrease in CPU utilization, on a CPU core with half the clock speed• Lead an optimization effort of the top-level control loop on Spot, for an overall 1.9x speedup

Waltham, Ma, Us

Compute Research Team (May 2022 - October 2022)• Research and experiment with the latest robotics algorithms and hardware, in order to accelerate existing applications, enable new abilities, reduce costs, and improve performanceStretch Platform Team (July 2020 - May 2022)• Deployed Wi-Fi on Stretch, involving: developing a hardware qualification test plan, experimenting with network architectures, rewriting a significant portion of the network stack, and standing up and managing a new wireless network in the lab• Optimized and debugged Wi-Fi performance, achieving significant speedup and uncovering application-layer performance bugs• Deployed, secured, and managed a CVAT (image labeling tool) instance to Google Compute Engine as part of a project to build a pipeline of labeled training data for box detection• Rearchitected the Stretch simulator to containerize it and build generic components that were adopted by Spot, simplifying the build workflow and enabling the ability to run sim tests in the cloud

Bedford, Massachusetts, Us

• Designed, built, and deployed a C++ performance-monitoring framework that has been adopted throughout the engineering organization, for a highly multi-threaded, soft real time system • Automated the framework from end-to-end such that performance data is collected on robots, uploaded to the cloud, analyzed and presented in daily reports using SQL and python, and then notifications of performance regressions are sent to interested parties using AWS Lambda, with no manual intervention• Wrote a mutex tracing utility that runs in real time and generates a report on mutex contention and inefficiency• Prototyped a visual odometry algorithm that used vSLAM observations to approximate the robot’s trajectory• Developed an algorithm that could reliably detect from limited sensor data when a robot has been trapped in an alley• Technologies used: C++17, Linux, Bash, Python, AWS, SQL

Manassas, Va, Us

• Developed efficient, multi-threaded computer vision applications for an autonomous vehicle• Wrote an image processing application to search radar data for moving objects and publish their location over DDS, using C++ and OpenCV• Built a Kalman Filter to track detected objects in subsequent radar data frames• Developed an API to actuate a pan-tilt camera to turn toward radar-detected objects and take pictures

Bethesda, Md, Us

• Supported the design team in developing optimized and lightweight designs• Performed non-linear static analysis to determine plastic strain with the intention of material rupture• Utilized modal analysis to determine natural frequencies and develop designs that avoid resonance• Analyzed interactions between multiple deformable bodies using non-linear contact analysis• Designed parts for harsh vibration environments by performing high-cycle fatigue analysis

Shelton, Connecticut, Us

• Utilized 3D CAD and FEA tools extensively for design optimization for high volume manufacturing• Created and developed new mechanism designs within Design for Manufacture constraints• Employed rapid-prototyping techniques such as 3D printing to quickly develop concepts• Worked closely with Quality Engineers to address design or manufacturing issues• Designed and developed new parts for novel mechanism designs and prototype products

M.S., Computer Science

B.S. Aerospace Engineering & Mechanical Engineering