Rishabh Verma

Education

New York University — Brooklyn, NY
Master of Science - Mechatronics and Robotics, GPA: 3.8/4.0
Sep 2023 - May 2025

Visvesvaraya National Institute of Technology — Nagpur, India
Bachelor of Technology - Mechanical Engineering, GPA: 8.14/10.0
July 2019 - May 2023

Experience

BotCrew

Robotics Engineer (Autonomy & Perception)
May 2025 - Jul 2026

• Developed NMPC Controller for Reeds-Shepp trajectories with gradient-based smoothing, deployed on a Nvidia Jetson Orin.
• Built bracket detection using YOLO26 + TensorRT, generating waypoints for autonomous robot alignment during solar module placement.
• Led development of a FastDDS-based middleware pipeline for real-time data exchange across distributed robot modules.

Robotics Engineering Intern (State Estimation & Mapping)
Feb 2025 - May 2025

• Designed a ROS-agnostic mapping engine, scaling to 5,000+ acre maps enabling deployment on the company’s largest test field sites, showcased in live investor demo.
• Fused ORB-SLAM3 and OpenVINS odometry via an Extended Kalman Filter, improving navigation robustness in dynamic environments.

Agile Robotics and Perception Laboratory

Graduate Research Assistant (Collision Avoidance)
Sep 2023 - Jan 2025

• Co-authored research on NMPC-based obstacle avoidance using adaptive control barrier functions derived from real-time risk estimates at ICRA 2025.
• Built a monocular depth-completion pipeline using DepthAnythingV2, calibrating relative depth against sparse RGB-D data to generate dense time-to-collision maps.
• Optimized quadrotor platform design via firmware improvements, achieved 26% increase in flight time.
• Built a multi-source odometry pipeline (OpenVINS + PX4 IMU/barometer) for robust navigation in GPS-denied environments.

Defence Research Development Organisation (DRDO)

Undergraduate Research Intern (Autonomy)
May 2022 - Dec 2022

• Built a navigation pipeline for urban environments in CARLA, validated across various weather conditions and multiple maps.
• Implemented PID control and A* path planning for precise vehicle navigation.
• Deployed YOLOv5 pedestrian detection, achieving collision-free navigation across crowded intersections.

Publications

• A. Saviolo, N. Picello, R. Verma, J. Mao, & G. Loianno.
  “Reactive Collision Avoidance for Safe Agile Navigation”, IEEE International Conference on Robotics and Automation (ICRA), 2025.

• R. Verma, J. Phillip, M. Kotambkar.
  “From Concept to Reality: An In-Depth Review of Autonomous Driving Strategies and Their Implementation”, International Conference on Mechanical Design and Manufacturing, 2023.

Key Projects

Monocular Depth Estimation

GitHub

• Implementation of DenseDepth with a ResNet34 encoder swapped in place of the original backbone, paired with a bilinear-upsampled decoder and skip connections.
• Achieved inference at 30 FPS on a Jetson Orin NX with MIPI camera input, validating real-world deployability.

Semi-Autonomous Urban Litter-Cleaning Robot

GitHub

• Built a mobile robot combining Raspberry Pi + Arduino to autonomously detect and collect litter using an ultrasonic-triggered MobileNetV2 detection pipeline.
• Coordinated sensor data, image capture, and robotic arm control over UART for closed-loop trash collection.

Robust Quadcopter Control in Windfield

GitHub

• Designed and assembled a quadrotor from scratch, with off-the-shelf motors/ESCs/props and fabricating custom 3D-printed mounts.
• Designed a control architecture to reduce drift, benchmarking PID, LQR, MPC, and Geometric controllers in simulation at 1-5 m/s.

Skills

Languages and Tools:
C, C++, Python, Git, Docker, MATLAB, Simulink, Arduino IDE, SolidWorks

Robotics:
ROS2, Gazebo, PX4, CARLA, PyTorch, TensorRT, CycloneDDS, FastDDS, Optimal Control, Model-Based Control, Visual Odometry, Camera Calibration, State Estimation, Lucas-Kanade Tracking