Geometry-Based Overtaking & Traction-Aware Speed Control
Team research project · MTSU · Fall 2026 – Present
An interpretable, training-free alternative to learning-based overtaking in multi-agent F1TENTH racing. On a two-student team under Dr. Vishwas N. Bedekar, I lead trajectory planning and traction control: Frenet-frame track utilities (arc length s, lateral offset d) and curvature-limited passing trajectories built from arcs and splines that diverge, reach a peak offset, and merge back onto the raceline. IMU and wheel-encoder data are fused into slip proxies — planned versus measured lateral acceleration — to scale target speed to inferred traction limits through high-curvature corners and overtakes. I also lead metrics, logging, and analysis, benchmarking against Pure Pursuit, hybrid Follow-the-Gap/Pure Pursuit, and fixed-side overtaking baselines on lap time, overtake success, and slip events.
Hybrid Autonomous Racing Stack — F1TENTH Platform
Undergraduate research · MTSU · Summer 2026
A hybrid ROS2 racing stack that couples Pure Pursuit raceline tracking with reactive Follow-the-Gap obstacle avoidance, arbitrated by a switcher node that hands control between the two on LiDAR obstacle detection. Speed-scaled lookahead and PD steering gains were tuned for stable tracking of a pre-computed optimal raceline on the Spielberg circuit in the F1TENTH Gym simulator (ROS2 Foxy, Docker). The reactive path implements LiDAR preprocessing and gap selection — a safety bubble around the nearest obstacle, largest-gap steering, and speed scaled to steering angle — for collision-free overtaking at speed.
Autonomous Melty-Brain Combat Robots
Honors thesis research · MTSU Honors College · 2025 – 2026
Autonomous and non-autonomous Melty-Brain combat robots in the 3 lb Beetle weight class, designed, built, and programmed for the National Robotics Challenge. A Melty-Brain robot translates while spinning its entire chassis as a weapon, so heading must be reconstructed on the fly from onboard sensing to steer at all. The accompanying Honors thesis, defended in Spring 2026, examines the embedded AI and control systems that make this tractable alongside the ethical frameworks and safety protocols that autonomous weaponized robots demand. The work won the Ultimate Battle Royale Award and was presented at the Southern Regional Honors Conference.
Robotics for Strawberry Farming
NSF-funded research · MTSU Engineering Technology
An NSF-funded project applying robotics to strawberry farming, where fruit is fragile, ripeness is uneven, and field conditions are unstructured. I fabricate components of the robotic arm that transports harvested strawberries, working on the mechanical side of a pipeline that has to handle produce without bruising it.