Giving autonomous vehicles an open-source boost

Oliver Dröge

•

Jan 20, 2021

Autonomous driving

Giving autonomous vehicles an open-source boost

Oliver Dröge

Sr Manager Product Marketing

Jan 20, 2021 • 6 min read

Fully automated driving (AD) remains an elusive golden nugget. As companies continuously seek solutions, it’s taking some time and effort for AD’s full potential to take off. To speed things up, TomTom has partnered with the Autoware Foundation – making our map technology and delivery service available to their open-source community to bring us all one step closer to full, reliable AD.

Around the world, companies are working to make the leap from partial automated driving (AD) to a full-fledged autonomous experience. Why not open things up and put our heads together? TomTom decided to do just this by joining the Autoware Foundation: a non-profit organization supporting open-source projects that accelerate the development of AD technology.

Accessible automated driving starts with accessible technology

Having an open-source platform makes it easier for developers, engineers, experts and other researchers to collaborate and share their findings. This not only benefits automakers, Tier 1s and other key industry players, but it also encourages innovation from various global AD communities.

The Autoware Foundation released open-source software that has been qualified to run on self-driving vehicles on public roads in Japan since 2017: Autoware.AI. It is the world’s first all-in-one open-source platform for AD, which first launched as a research and development platform for researchers, developers and students. Carmakers and OEMs have also come on board, using Autoware to continue improving, experimenting and integrating self-driving technologies into vehicles.

With consistent research and expanding collaborations, this open-source software is giving AD technology a big boost. Already, Autoware.AI is supported by the largest automated driving open-source community, with 2,300+ stars on GitHub and 500+ accounts on Slack. Gaining widespread recognition and success across such communities is important to continuously improve open-source technology, encouraging others to contribute their ideas too. Furthermore, Autoware’s Mobility-as-a-Service model helps automakers to create more personalized modes of transportation, as well as greener and more economical alternatives for carsharing and on-demand mobility services.

Besides stimulating AD production and collaboration, the Autoware Foundation also advances education –through its open-source communities and dedicated Autoware courses in five countries. Players such as the Toyota Research Institute (TRI-AD), the US Department of Transportation and the University of Tokyo have already been using and contributing to the Autoware platform.

How TomTom tech supports Autoware’s open-source platform

Successful automated driving cannot happen without proper implementation of location technology. For self-driving vehicles to travel to their final destinations efficiently and safely they need precise, detailed maps that are constantly updated. And to ensure AD systems have the latest maps, they need a specific delivery service that can be easily integrated into the vehicle.

At TomTom, we provide both: high-definition (HD) maps and the TomTom AutoStream delivery service, which efficiently streams the most recent and relevant HD map content to in-vehicle systems from the cloud.

TomTom joined the Autoware Foundation as a Premium and Steering Committee member, contributing our HD map data and AutoStream service. In this partnership, AutoStream directly delivers map updates to the relevant Autoware.AI modules – allowing users to scale and enhance their AD applications by accessing TomTom's global coverage of HD maps.

How Autoware gives automated driving a boost

So, exactly what features does Autoware’s open-source software offer? Supporting numerous vehicle platforms and available under the Apache 2.0 license, Autoware includes the following modules:

Localization: Achieved by 3D maps and SLAM algorithms in combination with GNSS and IMU sensors.

Detection: Implemented by cameras and LiDARs that use sensor fusion algorithms and deep neural networks. This helps vehicles detect obstacles and the surrounding traffic environment.

Prediction and planning: Driven by probabilistic robotics and rule-based systems that partially use deep neural networks.

Control: Defines the motion of the vehicle with a twist of velocity, curvature and angle.

Simulation: Provides a test environment for AD system features.

Collaborating for an automated future

Through testing and collaboration, the partnership between TomTom and the Autoware Foundation will continue to drive innovation forward. TomTom’s globally scalable map services and Autoware’s open-source platform are two of the many features that will make AD a reality in the future. Of course, this partnership is not complete without our key contributors that continue playing a huge role in the advent of automation.