About Me
Welcome 🙏 I am Anubhav Tripathi. I am currently working as a Project Associate at the Robert Bosch Centre for Cyber Physical Systems (RBCCPS), IISc Bangalore, under Prof. Shishir Kolathaya. I was previously affiliated with the Robotics Research Center (RRC), IIIT Hyderabad, where I completed my M.S., working on learning-based control for robotic systems. My research interests lie in reinforcement learning, imitation learning, vision-language-action (VLA) models, and robotics, with a focus on building intelligent systems that can perceive, learn, and act in complex environments. Outside the lab, I maintain an active lifestyle through gym workouts, swimming, yoga, and cricket, while also enjoying reading to continuously expand my perspective.
Publications
Here are some of my recent publications:
- Learning Value-Guided Control Barrier Functions for Safe Manipulation with a Franka Panda Robot – Accepted for presentation at IEEE CASE 2026 (Work-in-Progress Track). Developed a learning-based safety framework for whole-body obstacle avoidance using Value-Guided Control Barrier Functions (V-OCBF) on a 7-DOF Franka Panda manipulator with simulation and hardware validation.
- Imitation Learning-based Control of Brachiation Motion with Anthropomorphic Hands(ICCAS 2025, Incheon) – I develop intelligent control systems for high-DOF brachiation robots using imitation learning and trajectory optimization, enabling dynamic swing-and-grasp locomotion in simulation. Download PDF
Google Scholar: View my full publication list
Portfolio
Here are some of my recent projects:
Safe Robotic Manipulation using Learned Control Barrier Functions
Developed a learning-based safety framework for a 7-DOF Franka Panda manipulator using Value-Guided Control Barrier Functions (V-OCBF). The system learns a state-dependent safety function from simulation and enforces real-time safety using a CBF-QP controller. Demonstrated safe obstacle avoidance with both simulation and hardware validation.
VLA-PTZ: Embodied Visual Servoing via GRPO
Developed a Vision-Language-Action (VLA) framework for autonomous PTZ camera alignment. The system utilizes DINOv2 spatial embeddings and a Qwen-2.5 3B policy optimized via Group Relative Policy Optimization (GRPO). Developed for the Meta PyTorch OpenEnv Hackathon.
View on GitHubSpatial Understanding for Small Vision-Language Models
Adapted a 500M parameter Vision-Language Model (SmolVLM) to improve 3D depth reasoning and spatial formatting within strict 8GB VRAM limits. Engineered a data-centric task-aligned pipeline and applied parameter-efficient fine-tuning (LoRA) to attention layers, yielding a significant accuracy improvement on the zero-shot CV-Bench benchmark.
View on GitHubCustom GPT
Built an AI-powered conversational system using the Groq API, capable of delivering concise responses and detailed programming explanations with support for syntax-highlighted outputs.
View on GitHubNumber Plate Detection
Implemented a computer vision pipeline using OpenCV and EasyOCR to detect and extract alphanumeric text from vehicle license plates with bounding box localization.
View on GitHubReal-Time Ball Catching with XArm7
Designed a real-time robotic system using inverse kinematics and predictive control to track and intercept a moving ball using the XArm7 manipulator.
Spam Classifier
Developed a machine learning model to classify messages as spam or ham, integrated with an interactive web interface for real-time user input and prediction.
View on GitHubShipment Price Prediction
Designed a regression-based machine learning model to predict shipment pricing based on structured logistics features.
View on GitHubSensor Fault Detection
Developed a machine learning system to identify faulty sensors, improving robustness and reliability in sensor-driven environments.
View on GitHub6R Manipulator Trajectory Tracking
Performed kinematic modeling and simulation to enable accurate trajectory tracking of a ball using a 6R robotic manipulator.
Walking Robot (Jansen Mechanism)
Engineered and prototyped a walking robot with Jansen linkage, incorporating onboard sensors for temperature and smoke detection with emphasis on robust locomotion.
Substation Monitoring System
Developed a real-time monitoring system for electrical substations to enhance safety, fault detection, and operational efficiency.
Voltage Cutoff with Alarm
Designed a protection device that detects abnormal voltage conditions and triggers automatic cutoff with an alarm to safeguard electrical equipment.
Technical Blog
Practical Notes on Bringing Up the Franka Panda Robot in a Research Environment: From Real-Time Kernels to Python Control
A practical guide to setting up and controlling the Franka Emika Panda robot using Linux real-time kernels, optimized Ethernet configuration, and Python-based control.
Read BlogPitch to PyBullet: A Beginner's Tutorial Guide for gym-pybullet-drones
A beginner-friendly, cricket-themed tutorial to master drone simulation from scratch. Learn how to inspect drone states, configure a steady hover, navigate coordinate waypoints, and execute safe landings inside the PyBullet sandbox.
Read BlogResume
Education
Post Graduate Diploma in Applied Statistics (PGDAS) – Indian Statistical
Institute (ISI), Kolkata
2025 – 2026
Master of Science (MS) - International Institute of Information Technology,
Hyderabad (IIITH)
2023 - Dec 2025
Advisors:
Dr. Nagamanikandan Govindan
,
Dr. Harikumar Kandath
Bachelor of Technology (B.Tech) in Electrical Engineering - Shri Govindram Seksaria
Institute of Technology and Science (SGSITS), Indore
2019 - 2023
Secondary and Senior Secondary Education - Tiny Tots Public Sr. Secondary School, Sultanpur, U.P.
Experience
Advisor: Prof. Shishir N. Y. Kolathaya
Project Associate
April 2026 – Present
Research Intern
January 2026 – March 2026