Recently in autonomous cars Category

The AutoRally Platform

From Brian Goldfain at Georgia Tech

We're the AutoRally team from Georgia Tech and we're pushing autonomous driving to the extreme with our AutoRally robot. Our robot reaches speeds of 20mph driving fully autonomously using only onboard sensing and computing at our test track, often powersliding around turns. We built this ROS compatible robot in-house as a high performance testbed for control and perception research because there was nothing like it commercially available. The platform is designed to be safe, robust, and accessible for testing aggressive autonomous driving. autorally_platform_gallery.jpg autorally_platform_twowheels.jpg We initially chose to use ROS as our middle layer for the same reason so many other use ROS: so we didn't have to start from scratch. The node structure, messaging, existing sensor drivers, and visualization tools helped us get our system up and running quickly and focus on pushing the limits of autonomous driving. As the project grew, we found new students with ROS experience gained from tinkering, clubs, or classes. Instead of spending months familiarizing themselves with custom software, they arrive with an understanding of core concepts and vocabulary required to immediately contribute to the project.

We want other ROS users to play with our code, build their own AutoRally platforms, then come race against our robots. Check out the AutoRally platform details here: http://autorally.github.io and a video of recent research on the platform presented at ICRA2016 titled "Aggressive Driving with Model Predictive Path Integral Control":

skip to the 2 minute mark if you just want to see the the results.

All of out our core code, a Gazebo world, and build instructions are available on GitHub: [https://github.com/AutoRally/autorally](https://github.com/AutoRally/autorally)

Driverless Development Vehicle with ROS Interface

| No Comments | No TrackBacks
Choose either the Lincoln MKZ or Ford Fusion as a development vehicle.

lincolnmkz.png
Full control of
  • throttle
  • brakes
  • steering
  • shifting
  • turn signals
Read production sensor data such as
  • gyros
  • accelerometers
  • gps
  • wheel speeds
  • tire pressures

There are no visual indications that the production vehicle has been modified. All electronics and wiring are hidden.




Marvin is an autonomous car from Austin Robot Technology and the Department of Computer Science at The University of Texas at Austin. The modified 1999 Isuzu VehiCross competed in the 2007 DARPA Urban Challenge and was able to complete many of the difficult tasks presented to the vehicles, including merging, U-turns, intersections, and parking.

The team members for Marvin have a long history of contributing to open-source robotics software, including the Player project. Recently, Marvin team members have been porting their software to ROS. As part of this effort, they have setup the utexas-art-ros-pkg open-source code repository, which provides drivers and higher-level libraries for autonomous vehicles.

Like many Urban Challenge vehicles, Marvin has a Velodyne HDL lidar and Applanix Position and Orientation System for Land Vehicles (POS-LV). Drivers for both of these are available in the utexas-art-ros-pkg applanix package and velodyne stack, respectively. The velodyne stack also includes libraries for detecting obstacles and drive-able terrain, as well as tools for visualizing in rviz.

The Marvin team has also released an art_vehicle stack that provides the libraries that make Marvin go, including their navigation system. You can try it out with their simulator built on Stage.

marvin.JPG

Professor Peter Stone's group in the Department of Computer Science has been using Marvin to do multiagent research. You can learn about the algorithms used in the Urban Challenge in their paper, "Multiagent Interactions in Urban Driving". More recently, they have been doing research in "autonomous intersection management". This research is investigating a multiagent framework that can handle intersections for autonomous vehicles safely and efficiently. As you can see in the video above, these intersections for autonomous vehicles can handle far more vehicles than intersections designed for human-driven vehicles. For more information, you can watch a longer clip and read Kurt Dresner and Peter Stone's paper, "A Multiagent Approach to Autonomous Intersection Management"

Many people have contributed to the development of Marvin in the past. Current software development, including porting to ROS, is being led by Jack O'Quin and Dr. Michael Quinlan under the supervision of Professor Peter Stone.

Stanford_Junior.640w.jpg

Junior is the Stanford Racing team's autonomous car that most famously finished in a close second at the DARPA Urban Challenge. It successfully navigated a difficult urban environment that required obeying traffic rules, parking, passing and many other challenges of real-world driving.

Those of you familiar with Junior are probably saying, "Junior doesn't use ROS! It uses IPC!"

That's mostly true, but researchers have recently started using ROS-based perception libraries in Junior's obstacle classification system.

From the very start, one of the goals of ROS was to keep libraries small and separable so that you could use as little, or as much, as you want. In the case of the tiny i-Sobot, a developer was able to just use ROS's PS3 joystick driver. When frameworks get too large, they becomes much more difficult to integrate with other systems.

In the case of Junior, Alex Teichman was able to bring his image descriptor library for ROS onto Junior. He has been using this library, along with ROS point cloud libraries, to develop Junior's obstacle classification system. Other developers on the team will also be allowed to choose ROS for their programs where appropriate.

You can find out more about Alex's image descriptor library at ros.org/wiki/descriptors_2d.

Find this blog and more at planet.ros.org.


Monthly Archives

About this Archive

This page is an archive of recent entries in the autonomous cars category.

components is the next category.

Find recent content on the main index or look in the archives to find all content.