DRS Research

DRS Research2020-09-19T16:17:32+01:00

Research @ DRS

Our research spans legged locomotion, perception, navigation and manipulation. Here’s a closer look at some of the work going on in Dynamic Robot Systems.

Research Project Pages

  • Visual Inertial Legged Navigation System: Our VILENS system fuses legged (quadruped) robot sensing with camera images and inertial sensing to achieve reliable and low drift tracking.
  • LIDAR SLAM: Our LIDAR System, called AICP, can build large scale 3D models including loop closure and pose graph smoothing. The system requires only 2 CPUs of a low power computer such as those running on our ANYmal robot.
  • DARPA Robotics Challenge: This multi-year competition brings together collaborators from Europe (ETH, Flyability) and America (U. Nevada, SNC, Berkeley) to build a team of robots for exploring underground mines, caves and urban facilities.
  • Active Mapping: Research into next-best-view active inspection with our ANYmal robot.
  • Global LIDAR Localization: Our LIDAR Localisation system, called Efficient Segment Matching System (ESM). It uses deep learning to match point cloud segments while only requiring a CPU for run-time operation.

Videos and Open Source Software

We heavily use ROS in our research. Check out our GitHub account and our open source projects.

Research News

13Dec, 2019

Testing ANYmal at RACE

On Tuesday the 26th of November, Oxford Robotics Institute visited RACE to test out their ANYmal at the mobility benchmarking lanes. This is part of our RAIN collaboration to test the reliability of different [...]

28Oct, 2019

VILENS

VILENS - Visual-Inertial Odometry for Legged Robots VILENS (Visual Inertial Legged Navigation System) is a factor-graph based odometry algorithm for legged robots that fuses leg odometry, vision, and IMU data. This algorithm was designed by David [...]

26Feb, 2019

Global LIDAR Localization

Global LIDAR Localization Localization using LIDAR has advantages over visual localization. In particular LIDAR has a great degree of viewpoint and lighting invariance. It is however less informative. We have developed machine learning algorithms to [...]

Go to Top