About Me

II’m Srikumar Brundavanam, an autonomy engineer passionate about building real-world robotic systems that are safe, reliable, and deployable. I currently work at Oshkosh Corporation, where I contribute to the development and testing of L3 autonomous vehicle systems across localization, safety, hardware integration, and autonomy stack validation.

My work includes improving EKF-based localization for better on-road accuracy, designing a Request to Intervene (RTI) feature for safety-critical driver takeover scenarios, evaluating next-generation compute platforms against NVIDIA Jetson systems, and supporting vehicle remote operation and autonomy stack integration. I’ve also worked on scenario-based testing frameworks, localization and mapping KPIs, and object detection and avoidance algorithms for clearance-aware autonomous refuse cart collection.

I earned my Master’s in Mechanical Engineering from Carnegie Mellon University, focusing on robotics, controls, machine learning, SLAM, path planning, computer vision, and decision-making. I’m especially interested in end-to-end autonomy: taking ideas from algorithms and simulation into integrated, vehicle-tested systems

Before joining Oshkosh, I interned at Monarch Tractor, integrating software with hardware platforms like the Jetson Orin AGX, and developing ROS-to-CAN test tools for robust vehicle communication. My academic path has included contributions to CMU’s Computational Engineering and Robotics Lab (CERLAB), where I developed vision-based planning and collision avoidance systems for quadcopters using deep reinforcement learning.

My core interests lie in perception-aware control, real-time embedded systems, and robotics software integration. I enjoy tackling challenges where performance, safety, and scalability intersect in autonomous robotics platforms.

Previously, I worked with Professor Zhaodan Kong at UC Davis on UAV systems for wildfire air quality data collection and with the Laboratory for AI, Robotics, and Automation on a retractable-wing drone-kite hybrid. Other projects have included MPC-based multi-vehicle control, point-cloud based map estimation using neural networks, and advanced aircraft control systems—from tiltrotor designs to canard configurations aiming to redefine next-gen aviation.

Beyond engineering, I love cooking and exploring new cuisnes,binging movies and tv shows and cstaying active in my everyday life. I believe the best solutions come from blending creatvity and persistence. I’m always looking for ways to apply that mindset in the pursuit of building smarter, safer, and more sustainable systems.