Robotics

New collision avoidance exercise at Unibotics robot programming website: DWA

Learn the Dynamic Window Approach algorithm hands-on with Python and ROS2.

Deep Dive

Unibotics, the browser-based robot programming platform, has launched a new collision avoidance exercise centered on the Dynamic Window Approach (DWA) algorithm. This classic local navigation method, similar to Reid Simmons' CVM algorithm, is a key component of the widely-used Nav2 navigation stack from Open Navigation LLC. The exercise is now available in Python on Unibotics (unibotics.org), with C++ and direct ROS support planned soon. It leverages Gazebo simulation and ROS2, providing a realistic environment for learners to experiment with real-time obstacle avoidance.

The exercise was developed by José Miguel Jiménez during his internship at JdeRobot, a robotics research group. DWA works by evaluating velocity commands based on robot dynamics and obstacle proximity, making it ideal for teaching fundamental robotics concepts. This addition expands Unibotics' educational library, which already includes ROS2 navigation training and other webIDE tools. The platform aims to make robotics programming accessible from any browser, lowering the barrier for students and professionals to gain practical skills with industry-standard algorithms.

Key Points
  • New DWA collision avoidance exercise on Unibotics (unibotics.org) using Python and ROS2
  • Based on the Dynamic Window Approach algorithm, part of the Nav2 stack from Open Navigation LLC
  • Developed by intern José Miguel Jiménez at JdeRobot; C++ and direct ROS support coming soon

Why It Matters

Hands-on DWA practice in a browser makes real-world robotics skills accessible without hardware.