Hi, I'm
Bharath Rajiv Nair
Building real-time perception systems for autonomous robots. Specializing in 3D perception, sensor fusion, and multi-object tracking.
About
I'm a Robotics Software Engineer at ArcBest Technologies, where I design and deploy real-time perception systems for autonomous warehouse robots. My work spans 3D collision avoidance, sensor fusion, multi-object tracking, multi-modal sensor calibration, and teleoperation pipelines.
My interests lie broadly in robotic perception — from traditional 3D perception tasks like sensor fusion and multi-object tracking, to the latest advances in transformer architectures, vision-language models, and robot learning.
I'm currently pursuing my M.S. in Computer Science (AI specialization) at Georgia Tech, part-time. I hold an M.S. in Mechanical Engineering from Columbia University (Robotics and Control) and a B.E. from BITS Pilani. Previously, I interned at Amazon Robotics as an Applied Scientist and have published at ICRA and other robotics conferences.
Technical Stack
Journey
Georgia Tech
M.S. in Computer Science (AI)
Georgia Tech
M.S. in Computer Science (AI)
ArcBest Technologies
Robotics Software Engineer, Perception
ArcBest Technologies
Robotics Software Engineer, Perception
Columbia University
M.S. in Mechanical Engineering
Columbia University
M.S. in Mechanical Engineering
Amazon Robotics
Applied Scientist Intern
Amazon Robotics
Applied Scientist Intern
SUTD
Research Assistant, ROAR Lab
SUTD
Research Assistant, ROAR Lab
Purdue University
Visiting Research Student
Purdue University
Visiting Research Student
BITS Pilani
B.E. (Hons.) in Mechanical Engineering
BITS Pilani
B.E. (Hons.) in Mechanical Engineering
Education
Georgia Institute of Technology
M.S. in Computer Science
Artificial Intelligence specialization
Columbia University
M.S. in Mechanical Engineering
Robotics and Control specialization
BITS Pilani
B.E. (Hons.) in Mechanical Engineering
Experience
ArcBest Technologies
Robotics Software Engineer, Perception — R&D Lab
Autonomous Forklift Perception
3D perception pipeline for autonomous navigation and obstacle avoidance. Integrated comprehensive sensor suite including LiDARs, stereo cameras, and ToF sensors for warehouse operations.
Perception Scope & Strategy
Led perception project scoping and decision-making, aligning sensor and algorithm choices with customer requirements and operational constraints.
3D Safety & Scanning Systems
Developed real-time 3D collision avoidance using multi-LiDAR fusion (>98% success rate) and YOLOv8-based barcode scanning for semi-autonomous teleoperation.
Multi-Modal Sensor Calibration
Architected a modular, configuration-driven calibration framework (camera-camera, camera-LiDAR, LiDAR-LiDAR), enabling fully remote calibration by teleoperators.
LiDAR BEV Visualization
Built a real-time Bird's-Eye View visualization system with dynamic ROI filtering for improved spatial awareness during teleoperation.
Edge Deployment on NVIDIA Jetson
Integrated multi-modal perception sensors on Jetson Orin, building real-time pipelines using ROS and PCL. Deployed at the first two customer sites.
Amazon Robotics
Applied Scientist Intern, Virtual Systems
Multi-Agent Warehouse Simulation
Built a proof-of-concept simulation for realistic multi-agent warehouse motion — workers, pallet jacks, forklifts — praised by the team's software architect for producing the most realistic entity motion they had seen in simulation.
Path Planning & Obstacle Avoidance
Implemented multi-agent path planning, trajectory execution, and dynamic obstacle avoidance in Python, enabling scalable layout optimization and robot throughput analysis.
Singapore University of Technology and Design
Research Assistant, ROAR Lab (Prof. Mohan Rajesh Elara)
sTetro-C Reconfigurable Robot
Worked on a self-reconfigurable service robot for autonomous staircase navigation and cleaning. Designed the reconfigurable cleaning mechanism and platform.
Staircase Perception & Navigation
Selected and integrated sensors for staircase perception. Filed a design patent for the reconfigurable architecture.
Publication
Contributed to a peer-reviewed paper in Expert Systems with Applications on staircase navigation using a self-reconfigurable service robot.
Purdue University
Visiting Research Student, Collaborative Robotics Lab (Prof. Richard Voyles)
Neuromorphic Tactile Skin
Developed a neuromorphic neural network for tactile robotic skin in Python/Keras, optimizing printable, hardware-constrained analog designs.
ICRA 2021 Publication
Research published at IEEE International Conference on Robotics and Automation (ICRA) 2021.
Projects
SensorLens
An IDE for your tracker — interactive 3D visualization and debugging tool for multi-object tracking on autonomous driving datasets. Step through frames, inspect every bounding box, and see exactly where your tracker fails (ID switches, false positives, missed detections) with color-coded MOT diagnostics. Supports nuScenes, KITTI, Waymo, and Argoverse 2 through a universal dataset-agnostic scene format.
Publications
Embedded neuromorphic architecture for form+function 4D printing of robotic materials
IEEE ICRA 2021PaperTowards staircase navigation and maintenance using self-reconfigurable service robot
Expert Systems with Applications, 2025PaperCollaborative perception in multi-robot systems: case studies in household cleaning and warehouse operations
ICRCV 2024PaperAdvancing robotic surgery: affordable kinesthetic and tactile feedback solutions for endotrainers
CCRIS 2024PaperA modular and autonomous staircase-climbing robot with mopping module
Technical Disclosure, SUTD, 2020Get in Touch
I enjoy connecting with people who are passionate about robotics, perception, and intelligent systems. Whether it's a conversation about 3D sensing, a collaboration, or just a shared curiosity — I'd love to hear from you.