Loading
0
Quadruped Locomotion   SpaceData Inc.
SpaceData Inc.   Internship Project   2025

Teaching robots
to walk through
the wreckage

I trained a multi terrain locomotion policy for quadruped robots in Isaac Lab and tested it in indoor environments without GPS, built to simulate post disaster scenarios.

Isaac Lab
RL Training
ROS 2 Nav2
Navigation
Isaac Sim
Simulation env
Quadruped
Platform

When a building collapses, most survivors are found in the first 72 hours. Sending humans into unstable debris is dangerous, wheeled robots get stuck, and drones cannot carry sensors close to the ground. Legged robots can go where nothing else can.

The problem

Disaster environments
break every
assumption

After earthquakes, explosions, or structural failures, the operating conditions are not like anything a standard robot is designed for. Floors are uneven, corridors are blocked, there is no GPS, and lighting is unpredictable. Every second of delay matters.

Wheeled platforms lose traction on rubble and cannot climb over obstacles more than a few centimetres tall. Aerial robots have short endurance and cannot get close enough to debris piles to be useful. Legged robots are different. They can step over obstacles, keep balance on uneven ground, move through narrow passages, and carry sensor payloads deep into environments no human should enter.

The hard part is making them autonomous enough to do this reliably without constant operator input. That is what this project worked toward.

Operating conditions
Terrain
Debris and rubble
Irregular surfaces, step obstacles, narrow passages, unstable ground
Navigation
GPS denied
All localisation onboard using SLAM and sensor fusion
Visibility
Low and variable
Dust, smoke, damaged lighting, constrained sensor range
Time pressure
72 hour window
Most survivors found in the first three days after a structural collapse
Operator load
Minimal oversight
Emergency crews cannot dedicate attention to full teleoperation under field conditions
Platform
Quadruped robot
Four legged platform, step climbing, dynamic balancing, sensor payload capacity
The multi terrain policy I trained during this internship is under an NDA with SpaceData Inc. and cannot be shown publicly. The video below shows the flat terrain locomotion baseline that I extended the work from.
Flat terrain locomotion baseline   Isaac Sim   ROS 2 Nav2
What I built

Multi-terrain
locomotion
from scratch

At SpaceData Inc. my job was to extend a flat terrain locomotion policy so it could handle the kinds of surfaces a robot would really meet in a disaster zone. To do this I trained a reinforcement learning policy in Isaac Lab that keeps stable locomotion across rubble, slopes, steps, and irregular ground without pre mapping the terrain.

  • Trained quadruped locomotion policy using RL in Isaac Lab
  • Multi terrain curriculum across flat ground, rubble, slopes, and steps
  • Reward shaping for stability, smooth motion, and disturbance recovery
  • Policy outputs velocity commands compatible with ROS 2 Nav2
  • Navigation stack integrated with Isaac Sim for full autonomy pipeline
  • Disaster style indoor environments built in Isaac Sim for validation
  • Localisation without GPS using onboard SLAM
  • Deployment architecture designed to run on physical hardware
Status

Functional in sim.
Ongoing
towards hardware.

The multi terrain policy reaches its navigation goals reliably across the simulation environments I built. The next steps are to validate it on physical hardware, tighten the navigation integration for narrow corridors, and improve recovery behaviors after unexpected falls.

Most of the detailed results from this internship are under NDA. What I can say is that this was the first time I had to build something with a real operational constraint. It was not a research demo. It was a system someone would actually deploy in the field, and that changes how you think about every design decision.