Lunabotics @ Columbia
ATLAS is Columbia University's reentry into the NASA Lunabotics Challenge, a national engineering competition focused on autonomous lunar excavation. The goal: design a robot capable of traversing simulated terrain and constructing a berm using lunar regolith simulant.
This year marked the revival of Columbia's Lunabotics program after a multi-year hiatus. As co-leads, we chose to pivot from full competition entry to team formation, research, and foundational design, recognizing that building a strong technical and organizational base was essential before competing with legacy programs.
The result: a high-impact first year that produced a fully fabricated chassis, detailed wheel designs, material testing infrastructure, and a novel unbuilt core-drilling subsystem, while training a new generation of engineers to carry the project forward.
September 2024 – Present
We made a strategic decision: rather than rush toward a fragile competition entry, we focused on building a robust foundation including hardware prototyping, subsystem R&D, recruitment, and engineering education.
Outcome: Delivered designs, prototypes, documentation, and trained leadership for next year's build.
Designed and built a full-scale chassis via waterjet and break forming. Optimized geometry for debris clearance and modular integration with autonomy and power systems.
Status: Completed and assembled for static and interface testing.
Created two distinct wheel concepts:
Performed CAD, FEA, and early testing with regolith simulant.
Status: Fully designed and analyzed, fabrication pending.
Designed a barrel-style core drill that served both as an excavation and containment system for regolith. The mechanism is driven by twin worm gears, coupling translation and rotation while still allowing independent control of each motion. This enabled a compact, mechanically robust design suited for lunar excavation constraints.
Status: Fully designed but not fabricated.
Tested 3D-printed TPU anisotropy, performed abrasion tests with LM-1 lunar regolith simulant, and compared aluminum, titanium, and composites for structural use.
Status: Material selection matrix complete, findings handed off to future team.
Set up control architecture using Raspberry Pi with Ubuntu and ROS2, integrated LiDAR sensors, and successfully visualized data using RViz2. Began work on a Simulink digital twin for motion planning simulations.
Status: LiDAR visualization achieved, localization and navigation in progress.
We are proud of this year's pivot and progress. By laying a solid technical and organizational foundation, we have positioned Columbia to reenter the competition field with a mature, well-trained team ready to compete seriously in future Lunabotics challenges.