Dan Shafman

Bachelor's

Computer Science

CATERPILLAR

Autonomy Engineer

• Developed ROS-like middleware-agnostic (MWA) software framework to manage complex, concurrent positioning system • Designed and built Hardware-in-the-loop (HIL) simulation bench from ground up to expedite validation process • Created library to generate vehicle trajectory from fiducial markers using modern polyline-offsetting algorithm • Implemented feature to record and analyze runtime metrics from MWA system, for ease of testing • Created client-facing OpenGL tool to visualize fiducial markers in 3D environment around autonomous vehicle in real-time • Created multiple validation-assist tools and HUD elements using OpenGL and FLTK, and performed field testing with them • Scripted several data analysis tools using Bash to generate useful metrics on logged field data collections • Outlined in-depth regression test plan from scratch on Azure DevOps to validate team progress after every Agile sprint • Read research papers monthly to keep up with state-of-the-art algorithms and held seminars to present findings to team • Improved localization time and scope by suggesting and implementing new algorithm to replace inefficient system