ROS Resources: Documentation | Support | Discussion Forum | Service Status | Q&A

The Best Things for the Holidays are Free


The holidays are finally here. For many of us that means a nice solid break from work to eat food, play games, hang out with the family, and work on personal projects. If you are looking for something fun to do this holiday season, brush up on your ROS skills, and maybe make a few bucks, we’ve got a lot of exciting competitions using ROS right now. The best part is you don’t need to order a lot of parts or spend hours at a bench assembling a robot, you can do everything in Gazebo.

All the SubT Vehicles and Payloads

The first competition you should be aware of is the DARPA Subterranean challenge (SubT). This competition is the latest in a long line of robotics challenges like the DARPA Urban Challenge and the DARPA Grand Challenge. A number of recent events like the Thai cave rescue, a variety of mining disasters across the globe, and the flooding of subterranean transit lines due to global climate change, have underscored the need for search and rescue robots that can operate in confined spaces where GPS is not available. This competition seeks to push the state of the art in mobile robots that operate in dark, confined, and dangerous spaces. The challenge is divided into three individual competitions each in a different subterranean environment (tunnels, urban terrain, and caves). The first phase of competition wrapped up in August and you can read all about it in this blog post by DARPA.

SubT Virtual Track

The DARPA SubT challenge has both virtual and real competition components, with the virtual competitors sharing US$1.5M in prizes. Getting started is super easy and only requires a reasonably performant computer (we suggest something with a 4GB+ nVidia graphics card). To make things even easier, DARPA provides competitors a variety of pre-configured robots, a suite of sensor payloads, and a virtual practice arena. All you have to do is select your robot, select your sensor payload, and then start building your robot’s behavior. There is a pre-configured docker container that you can use to try out the different robots and explore the terrain. I was able to have a virtual robot moving a virtual track in a little over an hour. If you are interested in joining the competition head over to to get all of the details.

JPL Open Rover

If space travel is more your thing the Jet Propulsion Laboratory (JPL) in conjunction with the Amazon RoboMaker and SageMaker teams have put together the AWS JPL Open Source Rover Challenge. For this challenge you’ll use AWS RoboMaker to build a navigation algorithm to move a simulated Martian rover between waypoints on the Martian surface. What’s cool about this competition is that it is entirely in the cloud, so you really can participate with just a simple laptop and a decent internet competition. AWS and JPL have generously put up US$20,000 in prizes for the competition along with a number of AWS credits. Since the robot’s environment and hardware configuration is locked down by JPL the competition focuses on building a reinforcement algorithm to train the robot’s behavior making it a bit easier to get started. A full description of the technical details and sensor suite can be found on the competition’s GitHub repository.

ARIAC Simulation Challenge 2019 Screen Capture

The last challenge that is currently running is the National Institute for Science and Technology (NIST) Agile Robotics for Industrial Automation Challenge (ARIAC). This is the third year of the NIST ARIAC challenge, and this year’s challenge is an extension of the pick and place activities that were simulated in the previous years.This year there will be a US$17,500 prize pool split between the first, second, and third place competitors. You’ll need to register to get full competition details, but a summary of the previous year’s events is available on the project’s BitBucket account. Since the competition hasn’t quite started yet, the final details haven’t yet been released, but that doesn’t mean you can’t fire up last year’s simulations and see what you can make happen. ROS Discourse has a whole topic already dedicated to ARIAC so you can see what has worked for competitors in previous years.