BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Will A Small Open-Source Effort From Japan Disrupt The Autonomous Space ?

Following
This article is more than 3 years old.

The article  “Linux Beat IBM, will open-source software beat Waymo and Tesla ?” introduced the  power of the open-source organizational model and explored its application in the space of autonomous systems. “Open Source AVs: The story of AV Development in Estonia,”  detailed a project by the Tallinn University of Technology (TalTech) which built a functioning AV Shuttle in only a year using open-source infrastructure. What did Taltech use to achieve this impressive result ?  An open-source solution called Autoware which was built by a small team in Japan.  

What is Autoware ?

In August 2015, Professor Shinpei Kato from Nagoya University released the initial release of an open source system called Autoware.  Autoware is an operating system for autonomous vehicles which contains all the major components (localization, mapping, route planning, vehicle/pedestrian identification, tracking, simulation, lane detection, sensor drivers/fusion, etc). 

“I started the Autoware project in order to enable as many individuals and organizations as possible to contribute to open innovations in autonomous driving technology. With a baseline reference design, we wanted to provide (i) an open-source software platform for production, education, and R&D and (ii) business opportunities and schemes using the open-source software platform,”  said Dr. Kato.  

Soon after initial release, the industry demand was too intense to be handled by an academic research group, so in December 2015, Tier IV, a private company, was formed to handle the maintenance and support load for Autoware.  As the interest for collaboration grew and governance issues became more relevant, Tier IV transferred its assets to The Autoware Foundation (nonprofit) in December 2018 in an open-source style structure. Autoware adopted a fairly standard Apache 2.0 open-source licensing structure.   

The Autoware Foundation boasts over 50 members with a mix from semiconductors(NXP), electronic device automation (Cadence), sensor companies(Velodyne),  and academia(TalTech). A very interesting recent member was the USDOT Federal Highway Administration. The technical steering committee is organized by member companies, while the board of directors committee is in charge of the foundation management.

In terms of contributor momentum, the foundational members of the Autoware Foundation (Tier IV, Apex.AI, and Linaro) drive most of the activity with component suppliers (sensors) engaging at the device driver level. Dealing with the variety of contributors, while maintaining a coherent software system, is one of the ongoing challenges for any open-source project.  Mature open-source projects devote a significant amount of their resources on processes as well as automation for validation for incremental contributions. Autoware is making its way up this maturity curve on these topics.

“In terms of project momentum, we have hundreds of projects which are using our system to build autonomous solutions,” said Dr. Kato. One of the most visible is the Tokyo Olympics Shuttle(e-Pallette). Toyota is working with Tier IV to provide an automation solution with the e-Pallette solution for the Tokyo Olympics, whenever they happen.  Toyota boasts deals with Amazon AMZN , Didi Chuxing, Mazda, Pizza Hut, and Uber on the e-Pallette mobility concept.

Overall, in the space of Autonomous Vehicles, the vast majority of development is done inside private companies such as Waymo and Tesla with proprietary solutions. Even the “open” solutions which are available in the marketplace are dominated by large companies such as Baidu BIDU and Nvidia NVDA . Autoware is a very interesting independent open-source effort which is gaining momentum. As an indication of the growing maturity of the project,  they have established Autoware.Auto (production quality certifiable solution) and Autoware.IO (interface definition for third-party software).   

With Linux, a small team in Finland built sufficient momentum to change the center-of-mass in the corporate computing space. Perhaps the next big shift will come from a small team in Japan in the autonomous vehicle space.  

Follow me on Twitter or LinkedInCheck out my website