Antar Mazumder
Machine Learning Engineer | Robotics Generalist | Researcher | Mechatronics Engineering Graduate
Machine Learning Engineer | Robotics Generalist | Researcher | Mechatronics Engineering Graduate
Hi! I’m a PhD student in Robotics, specializing in Space Robotics, at the Colorado School of Mines. I hold a BSc in Mechatronics Engineering from RUET and previously worked as a Machine Learning Engineer, primarily in computer vision. My research focuses on robot perception, digital twins, and enabling robots to navigate extreme, unstructured environments—think Martian lava tubes. I also explore human-robot interaction and robotic exoskeletons. Currently, I lead the Digital Twin team at the Robotic Space Exploration (RoSE) Lab, working on the Mines Lunar Surface Simulation (MLSS) Testbed.
Some highlights of my work include a six-wheeled rover simulation for 3D mapping Martian caves, featured in ROS News (Jan 2022), and RTG, a monocular depth estimation-based terrain generator, also featured in ROS News (Feb 2025). I’ve published research on digital twins in robotics and off-world navigation, and I led Team Ogrodoot to rank 11th globally in the International Rover Design Challenge 2023. When I’m not debugging robots or building machines to explore Mars and the Moon, I’m geeking out over open-source robotics, dreaming up space missions, or tinkering with ROS packages late at night.
“Our deeds determine us, as much as we determine our deeds.”
-George Eliot
Recent Activities
Presented RoSE Terrain Generator and other Works at Machine-Ground Interaction Consortium (MaGIC) 2025
New DEM reconstructed of with craters in Mines Lunar Surface Simulation (MLSS) Testbed