Autoware is an “All-in-One” Open-source Software for Autonomous Driving

Orange Pi Development Boards

All major automotive companies, and some technology companies are all working on autonomous driving with the ultimate goal of achieving level 5 autonomous driving meaning no human intervention is needed at any stage.

Development will take some more time, and companies are now competing with closed source software and hardware. But as I browsed through Linaro Connect Bangkok 2019 schedule, I found out there’s an open source autonomous driving software called Autoware.AI.

Autoware.AI Software Architecture
Autoware.AI Software Architecture – Click to Enlarge

Several “Autoware” projects are managed by the newly founded Autoware Foundation, a non-profit organization created to develop a synergy between corporate development and academic research in order to provide access to autonomous driving technology for everyone:

  • Autoware.AI is the first version built on ROS 1, and Linux, and has been developed as a research and development platform
  • Autoware.auto is the second version built on ROS 2, and Linux, with a complete redesign.
  • Autoware.IO is an interface project for Autoware to be extended with proprietary software and third-party libraries for example drivers for sensors, by-wire controllers for vehicles, and hardware-dependent programs for SoC boards.

Autoware supports Localization, Mapping, Object Detection & Tracking, Traffic Light Recognition, Mission & Motion Planning, Trajectory Generation, Lane Detection & Selection, Vehicle Control, Sensor Fusion, Cameras, LiDARs, RADARs, Deep Learning, Rule-based System, Connected Navigation, Logging, and more. Autoware is best suited for urban cities, but highways, freeways, mountainous regions, and geofenced areas can also be supported.

Autoware Simulator Demo
Click to Enlarge

While in theory, you could use Autoware with actual autonomous vehicles, you may want to get started with something a bit more safe: a ROSBAG-based simulator that allows you to try it out with demo data. You’ll need a fairly powerful computer for the simulation as an 8-core Intel processor combined with 32GB RAM and an NVIDIA Geforce 980M GPU or greater is recommended. Alternatively, it’s also possible to run Autoware on NVIDIA DRIVE PX2 / Xavier, or Linaro “Syncquacer” Developer Box, but the latter is not recommended since it lacks performance and some features are not supported.

Autoware is released under an Apache 2.0 license with source code and documentation available on Github. You may also want to visit Autoware.org for more details. The project may end-up in actual vehicles as companies like Arm, Linaro, LG, Huawei, SiFive, Intel, Xilinx, and TRI-AD are all members of the Autoware Foundation. I could not see any car manufacturers directly involved, but TRI-AD stands for “Toyota Research Institute Advanced Development Inc.”, a new company (July 2018) established by Toyota Motor, Aisin Seiki and Denso with the goal of hiring about 1,000 staff working on “fully-integrated, production-quality software for automated driving”.

Support CNX Software - Donate via PayPal or become a Patron on Patreon

5
Leave a Reply

avatar
1 Comment threads
4 Thread replies
0 Followers
 
Most reacted comment
Hottest comment thread
3 Comment authors
JohanneschaxSander Recent comment authors
  Subscribe  
newest oldest most voted
Notify of
Sander
Guest
Sander

OK, nice. But what is the status of George Hotz’s $1,000 self-driving car kit … that became https://comma.ai/

A first add-on for my modest Volkswagen would be traffic sign recognition. My SO’s Volkswagen (a Passat) has that built in for speed traffic sign … quite handy.
Oh, and changing my cruise control into Adaptive cruise control would be fantastisc. I can recommend Adaptive cruise control to everybody.

chax
Guest
chax

AFAIK the comma.ai project now consists of various hardware devices and an open source software platform. The devices are a mobile phone for camera view and some kind of plug in module to connect to the OBD2 port of the vehicle. Some US Honda’s and Toyota’s seem to be able to control (steering and adaptive cruise control) with those devices but it is all very (car) hardware specific of course.

Maybe tapping into the throttle sensor combined with a mobile phone for traffic sign recognition or car distance recognition would do but it’s not really safe to mess with by yourself I guess ;).

Or you could try some app like https://play.google.com/store/apps/details?id=com.evotegra.aCoDriver&hl=en_US

Sander
Guest
Sander

There’s an app for that? Nice! I see “aCoDriver 5” has low reviews, but I’ve installed it anyway.

Sander
Guest
Sander

I installed the app aCoDriver, used it, and removed it: it did not work at all.

Johannes
Guest
Johannes

It’s an entirely different league. Comma.ai is a low end system for cruise control mostly based on cellphone hardware using (single) cameras. This system here requires vastly more expensive hardware, but should be able to drive in more complicated scenarios, using environment maps and more sensors like lidar laser scanners.
Not sure Autoware will really go into production. The project came from a university project, and they founded a company based on it.