Joshua Chong

Guiding the path planning group on CMU Racing's driverless stack, where planning, localization, and simulation work have to come together under race-day constraints.

Part of CMU Racing's Formula SAE Electric Michigan 2026 finish: 2nd in Driverless and 4th overall among 100+ teams.

• Built a B-spline planner around minimum-curvature paths, cutting simulated lap time by 17% in CarMaker runs.
• Designed localization logic that blends wheel odometry with cone observations through an Extended Kalman Filter.
• Prototyped graph-based SLAM in ROS 2 with GTSAM/iSAM2, including association checks, loop-closure handling, and offline review tools.
• Helped coordinate planning work inside a 150-person engineering team through design notes, reviews, issue tracking, and sprint planning.

Owned much of the robot autonomy code across several VEX seasons, tuning path following and control behavior as the game and hardware changed.

Competed at VEX Worlds in 2022, 2024, and 2025, with a global top-1% ranking and repeated state qualifications.

• Developed and tuned autonomous path following, odometry, and control logic across multiple VEX seasons under changing game constraints.
• Built and refined PID, odometry, and adaptive pure pursuit behaviors to improve robot reliability and match execution in competition.
• Supported team strategy and execution through repeated state qualification and qualification for VEX Worlds in 2022, 2024, and 2025.

Studied how immersive workplaces can be made safer and more inclusive, turning a broad literature review into conference-ready research.

Co-authored work accepted to IHSI 2025 on ethical and safety questions in VR/AR work environments.

• Synthesized findings from 300+ papers to surface recurring risks around access, safety, and workplace design.