Kshitij Madhav Bhat

Kshitij Madhav Bhat

MS Robotic Systems Development (MRSD) Student

Robotics Institute, CMU

About Me

I’m a first-year MS student in Robotic Systems Development at the Robotics Institute, Carnegie Mellon University.

I can be reached at kmadhavb@cs.cmu.edu

Education
  • MS Robotic Systems Development, Aug. 2024 - May 2026

    Carnegie Mellon University

  • B.Tech. Mechanical Engineering, Nov. 2020 - July 2024

    Indian Institute of Technology, Indore

Experience

 
 
 
 
 
IIT Delhi
Research Intern
August 2023 – December 2023 Remote - New Delhi, India
• Assisted in the design of a generative graph neural network enhanced with topological regularization, to improve SLAM for autonomous driving by removing dynamic objects from sparse LiDAR point clouds.
• Demonstrated superior performance against existing methods (44% lower Chamfer Distance than state-of-the-art) in three real-world and simulated datasets, against five distance metrics with 32 times sparse LiDAR scans.
• Contributed to a deep generative model for adversarial point injections on LiDAR scans, demonstrating superior performance in degrading map quality without compromising scan integrity on KITTI and CARLA-64 datasets.
 
 
 
 
 
Lakehead University
Research Intern
May 2023 – August 2023 Thunder Bay, Ontario, Canada
• Engineered a four-wheel steering and driving (4WS) mobile robot platform using Design for Assembly (DFA) techniques, resulting in a modular and easily maintainable platform for kinodynamic navigation algorithm development.
• Developed C firmware for Raspberry Pi Pico microcontrollers, and integrated PID control to actuators and enabling joint state data collection and velocity feedback, enhancing navigational accuracy in unstructured environments.
• Created hardware-agnostic C++ software to interface PID angle and velocity controllers using the ROS Control framework via serial communication, leading to general middleware suitable to any 4WS robot platform.
 
 
 
 
 
Ati Motors
Summer Intern, Autonomy Software
May 2022 – July 2022 Bengaluru, India
• Optimized Model Predictive Control (MPC) for Autonomous Mobile Robots (AMR) to achieve a 3x reduction in the turning radius, enabling sharp turns and in-place manoeuvres for space-constrained industrial units.
• Reformed MPC cost function optimization problem to consider physical motor constraints and latency, redesigned Jacobian matrix to speed up the solution compute time and validated it with comprehensive on-site testing.
• Shaped a post-processing routine for raw point cloud and IMU data in Rosbags for easy integration and validation of state-of-the-art 3D-LiDAR-based SLAM algorithms on real-world datasets of warehouses and shop floors.

Other publications

(2024). SLACK: Attacking LiDAR-Based SLAM with Adversarial Point Injections. IEEE ICIP Challenges and Workshops (ICIPCW) 2024.

PDF Cite DOI

(2022). Understanding the Heat Transfer Characteristics and Axis Switching Phenomenon in High Aspect Ratio Elliptical Orifice Impinging Synthetic Jets. Fluid Mechanics and Fluid Power, Volume 1.

PDF Cite DOI