
Hi, I am
Pavan Varma Pothuri
Full Name : Jagadeswara Pavan Kumar Varma Pothuri

MS in Robotics with a passion for autonomous mobile robots, machine learning, computer vision, and real-world problem solving. I design intelligent systems that perceive their environment, learn from data, and make decisions to navigate and act with purpose and precision.
About Me
I’m a robotics engineer passionate about designing intelligent systems that solve real world problems. With a Master’s in Robotics from the University at Buffalo, I’ve built a strong foundation in autonomous systems and artificial intelligence, grounded in hands on experimentation and real world deployment.
I thrive at the intersection of software and hardware where code becomes motion and algorithms must adapt to dynamic, uncertain environments. My experience has shaped my ability to think both analytically and creatively, translating complex concepts into robust, reliable systems.
I’m driven by the opportunity to learn, to collaborate with purpose driven teams, and to build technologies that make a tangible impact. To me, great engineering is about more than just precision. It is also about curiosity, persistence, and a deep understanding of how systems behave beyond the lab.
At my core, I enjoy solving hard problems, building systems that move, and continuously growing through each project, conversation, and challenge I take on.
Robotics Specialist
I specialize in robotics with a focus on designing and deploying intelligent systems that integrate perception, decision-making, and control. My work spans autonomous navigation, reinforcement learning, and multi-agent coordination, with hands-on experience in both simulation and real-world UAV platforms.
I’ve contributed to research publications and technical projects that leverage machine learning and optimization to build adaptive, reliable robotic systems capable of performing in complex, dynamic environments.

My Skills
Specialized technical expertise in robotics, AI, and autonomous systems development
Programming
- Python
- C++
- C
- MATLAB
- Bash
- CUDA
Robotics
- ROS & ROS2
- PX4 Autopilot, MAVROS, PyMAVLink
- SLAM
- Path Planning
- Trajectory Optimization, MPC, PID
- Multi-Robot Coordination
- Kalman Filter, EKF, UKF, Particle Filter
- HITL, SITL, Sim2Real Transfer
Computer Vision
- Object Detection
- Object Tracking
- Segmentation
- Optical Flow, Depth Estimation
- Camera Calibration, Stereo Vision
- Stereo Visual Odometry
- Structure from Motion (SfM)
- 3D Reconstruction
- Neural Radiance Fields (NeRF)
- Multi-View Geometry, Sensor Fusion
AI & ML
- Supervised & Unsupervised Learning
- Deep Learning
- Reinforcement Learning
- Transfer Learning, Multi-Task Learning
- Behavioral Cloning, Reward Modeling
- Diffusion Models
- Transformers
GenAI, LLMs, VLMs
- GPT, Gemini, Claude, LLaMA
- Vision-Language Models
- VLA Models for Action Generation
- Retrieval-Augmented Generation (RAG)
- LLM Fine-Tuning & Deployment
- LangChain, LlamaIndex
- AI Agents
Simulation Tools
- AirSim
- Gazebo
- Isaac Sim
- Unreal Engine
- PyBullet
- MuJoCo
Hardware
- Jetson Nano
- Raspberry Pi
- Pixhawk
- Crazyflie
- Parrot Anafi
- Intel RealSense
- LiDAR
- Arduino
Development Tools
- PyTorch, TensorFlow
- Keras, HuggingFace
- Scikit-learn, OpenCV
- Docker, Kubernetes
- Git, GitHub
- CI/CD, MLflow
- VS Code, Jupyter
- PostgreSQL, MongoDB
- Firebase, Slack
- Jira, Notion
Soft Skills
- Problem Solving
- Critical Thinking
- Team Collaboration
- Leadership
- Mentoring
- Knowledge Sharing
- Adaptability
- Time Management
- Continuous Learning
- Curiosity
Masters' Thesis
Algorithms and Physical Experiments for Vision-based Tracking and Operations of Unmanned Aerial Vehicles

Autonomous UAV Tracking
Developed an integrated tracking system for a chaser UAV to autonomously follow a moving aerial target using a hybrid approach that combines YOLOv11 for detection, KCF for temporal tracking, and a Kalman filter for velocity-aided state prediction. A PPO-based reinforcement learning controller was trained in simulation and deployed on a Crazyflie drone, demonstrating robust performance in both AirSim and real-world flight tests.
“A real-time vision-based UAV tracking system with physics-informed RL policies and successful sim-to-real deployment.”

Scalable and Load-Balanced Coverage Path Planning (SCoPP)
SCoPP is a scalable framework for coordinating multiple UAVs to cover large environments efficiently. It combines global Voronoi decomposition with local adaptive routing to ensure both spatial coverage and load balancing across drones. The algorithm was tested in simulation and deployed at the SOAR field site using Crazyflie and Parrot Anafi drones.
“An efficient multi-agent coverage strategy combining geometric partitioning with decentralized planning for scalable UAV missions.”
Professional Certifications
Continuous learning and professional development in robotics, AI, and autonomous systems
Let's collaborate
on Robotics & AI!
I specialize in building intelligent and efficient systems at the intersection of Robotics and Artificial Intelligence. Whether you're looking for a dedicated full-time engineer or want to team up on cutting-edge projects, I'm ready to connect.
I'm actively seeking full-time opportunities in Robotics and AI, and I’m also open to collaborative projects, consulting, and research partnerships. Let's innovate together, reach out anytime.
- +1 716 247 3865
- jpkvarmapothuri@gmail.com
-
35 Custer St
Buffalo, NY 14214