Nishit Mittal
I breathe life into machines by blending perception, control, and a touch of intelligence into their core.
Passionate about creating technology that bridges the gap between the digital and physical world, turning ideas into real-world robotics solutions.
To me, robotics isn’t just about engineering — it’s about infusing machines with purpose and autonomy.
Passionate about creating technology that bridges the gap between the digital and physical world, turning ideas into real-world robotics solutions.
To me, robotics isn’t just about engineering — it’s about infusing machines with purpose and autonomy.
Internships & Research Experience
The University of Queensland, Brisbane, Australia (QS World Ranking 2026: #42)
Visiting Research Student
Under Prof. Pauline Pounds (Professor, School of Electrical Engineering and Computer Science)
Under Prof. Pauline Pounds (Professor, School of Electrical Engineering and Computer Science)
(June 6, 2025 – July 31, 2025)
- Developed ground sensing and leg trajectory systems for the Tiny Giant Robot (TGR) bipedal walking robot.
- Contributed to work on terrain perception, locomotion control and stair/step handling.
- During our time there, we got TGR confidently walking over tricky terrains, steps, and curved slopes.
- The same robot was later showcased at the University of Queensland Open Day event on 3rd August 2025, where it successfully walked for hours over various terrains.
- A research publication from this work is anticipated.
- Received a Letter of Appreciation from Prof. Pauline Pounds in recognition of my contributions at the Robotics Design Lab.
Tiny Giant Robot
Orangewood Labs, San Francisco (Remote)
Robotics Software Engineer Intern
(August 6, 2024 – November 23, 2024)
- Converted an RL environment-based Doosan robotic arm simulation from ROS2 Foxy to ROS1 Noetic adapting for company's OWL robotic arm.
- Created a data collection script and modified launch files, improving backend data collection and model training.
- Worked on testing ROS and RealSense packages on Docker, implementing OpenVLA to some extent.
- Compared performance of visual models to optimize perception-based tasks.
- Additionally, documented the entire implementation process for each task separately.
Doosan Robotic Arm
Orangewood Robotic Arm
Visual Model Comparison (Speaker Detection)
Methodology
- Pre-trained on unlabeled speaker images using Lightly_Train.
- Used SimCLR for self-supervised pre-training (effective on small datasets).
- Fine-tuned the pre-trained model on labeled data using YOLOv8.
Observation
- Lightly_Train gives better accuracy, even with lower confidence than YOLOv8.
- Produces more stable and consistent bounding boxes.
- Overall detection is more reliable and precise.
Indian Institute of Technology Kanpur (IITK)
Research Internship
Under Dr. Tushar Sandhan (Assistant Professor, Dept. of Electrical Engineering)
(June 3, 2024 – July 31, 2024)
Topic: Monocular Visual Perception for Nano Drones
- Developed and implemented algorithms for enhanced monocular vision-based tasks on Tello Drone.
- Created 3D environment world in Gazebo and utilized Iris drone with MAVROS for simulation.
- Implemented pure image processing techniques over deep learning, optimized for computational constrained nano drones.
- Designed a vision-based surveillance application where the drone first detects a person, descends to avoid eye contact, and either (i) locates a flat wall surface to land and hide or (ii) detects and navigates through a window before landing.
Task 1: Autonomous Landing Near the Wall
Task 2: Passing Through a Window
Algorithm Validation
Familiarisation with the Tello Drone
Teleoperation with Keyboard
Hand Gesture Control
Face Tracking
Bharat Intern
Virtual Internship
(September 10, 2023 – October 10, 2023)
- Explored AI/ML tools like TensorFlow, PyTorch, and Scikit-learn.
- Conducted thorough data analysis (EDA) to understand datasets better.
- Worked on three different datasets from various domains.
- Achieved notable accuracy rates through machine learning techniques.
Major Achievements
International Rover Challenge 2025
Led 30+ member team as Vice-Captain to 1st Runner-Up position, outperforming all IITs and MSU Rover team from Russia.
Electronics & Robotics Club Hackathon 2024
Secured 1st position for developing and simulating an Autonomous Electric Failure Detection and Repair Robot.
Smart India Hackathon 2023
Created real-time monitoring dashboard for construction site using IoT devices and AI/ML tools, won ₹1,00,000 cash prize.
Vice Captain, Alakananda Martian Rover (IRC 2025)
Maintenance Task
- Responsible for precise and reliable control of the robotic arm during the competition. (Secured 3rd highest score in this task.)
- Operated the robotic arm to perform complex manipulation tasks including opening drawers, lifting objects such as cones and toolboxes, and interacting with knobs and switches.
- Interfaced and controlled the rover using a joystick; optimized the joystick control system for maximum responsiveness and accuracy.
- Refined control code to enable smoother joystick-based control for both the robotic arm and rover base.
- Contributed to the conversion of the control code from ROS1 to ROS2 this year, enabling improved modularity and real-time performance.
- Utilised SSH (Secure Shell) for command transmission to Intel NUC.
Automation Task
- Utilized an Intel RealSense depth camera and a YOLOv5 model, and made the rover turn automatically when it detected an arrow's direction.
- Developed and implemented visual alignment logic to ensure the rover accurately aligned itself with the arrow before initiating a 90 degree turn.
Astrobiology Task
- Developed and integrated the complete control system for the soil mechanical subsystem, enabling smooth operation via keyboard commands. (Secured highest points in this task.)
- The control system handled multiple components including the auger, drill, sample dropping mechanism, tray rotation as well as forward-backward, and peristaltic pumps for performing chemical tests.
- Implemented image stitching code to combine multiple site images into a wide-angle panoramic view, enhancing site documentation.
- Assisted in interfacing and retrieving data from key sensors such as GPS, IMU, NPK, soil moisture, and temperature sensors, contributing to the environmental analysis.
Communication System of the Rover
- Designed and led the development of the rover's communication system, ensuring reliable data transfer between the rover and the base station.
- Developed an HTML-based web interface using the Flask Python framework to enable real-time video streaming from multiple webcams mounted on the rover.
- Reduced latency significantly, allowing operators at the base station to control and monitor the rover with minimal lag, ensuring smooth task execution during the competition.
Projects
Robo Muse -Autonomous Mobile Robot
Implemented SLAM and obstacle avoidance on a mobile robot using LiDAR, encoder feedback, and EKF-based sensor fusion.
Orient - Ball Balancing Robot
Developed and programmed a custom 3-RRS manipulator, using Raspberry Pi 5 for real-time control with OpenCV, Inverse Kinematics and PID.
Gesture-Controlled Robot Car
Created an interactive and dynamic robot car responsive to hand gestures for movement control using Mediapipe and HC-05 bluetooth module.
Professional Certifications
NVIDIA Self-Paced Courses
Robotron Workshop Certificate
Thapar Summer School
Thapar ELC Certificates
Contact & Online Presence
