Key Highlights

Research Showcase: Work of many CoE representatives was highlighted in the Autoware Foundation social media channels to bring visibility and recognition to work.

Roadmap Taskforce: Carried over the conversation on the work that has been done at the Roadmap task force to inform the CoE members.

Off-Road Autonomy: Progress on datasets, training, and simulation for terrestrial and space exploration use cases.

Collaboration Opportunities: Upcoming EU projects, NASA/Space ROS alignment, and discussions on dataset standards and affordable platforms.

Autoware CoE Research Showcase

This month’s meeting highlighted a range of recent academic contributions from CoE members:

• Personalized Autonomy: Dr. Ziran Wang presented ongoing work using LLMs and VLMs to make behavioral planners more transparent and understandable to drivers.
• Vehicle Dynamics: Associate Prof. Hormoz Marzbani shared progress on scalable tire dynamics modeling and learning for control, supporting safer and more adaptive driving behavior.
• Automated Bug Detection: The UC Irvine team (Josh Garcia and Alfred Chen) demonstrated tools for automated bug finding and fixing, applied not only to Autoware but also to Apollo and OpenPilot. Their efforts led to NSF CAREER Awards for both professors.
• Clemson AutoDrive Platform: Venkat and his students showcased results from their simulation racing leagues with 60–70 teams, which ran without software crashes. Their digital twin environments are now being extended beyond racing to support off-road autonomy research.

Autoware Roadmap Taskforce Overview

The CoE meeting also reviewed progress from the Autoware Roadmap Taskforce, which launched earlier this year to coordinate execution across working groups. The taskforce is structured into four phases (P1–P4, three months each) and nine working groups, spanning core AV development, enabling technologies, and production/validation.

Three main objectives guide the effort:

• Incorporate cutting-edge AI-first technologies as Autoware transitions from v1.0 to v2.0.
• Deliver a production-ready stack that members can commercialize.
• Expand deployments of Autoware to more vehicles and increase autonomous mileage.

This structured approach ensures that research advances from CoE members can connect directly to Autoware’s roadmap, supporting the transition from lab prototypes to production deployments.

Off-Road Working Group Focus

The Off-Road WG, led by Po-Jen Wang, is taking an end-to-end approach to autonomy in extreme environments, spanning both terrestrial and space applications.

• Terrestrial vehicles: Off-road racing and mining vehicles, where LiDAR-based perception can be applied.
• Space exploration: Mars rovers and Lunar Terrain Vehicles, where harsh conditions make LiDAR impractical, requiring vision-based perception pipelines.

A modified version of the AutoSeg foundation model is being applied, trained on off-road datasets to deliver three key perception capabilities:

• Free space detection (drivable vs. non-drivable areas)
• Object detection (structures, vehicles, vulnerable living beings, etc.)
• Terrain classification (paved, dirt, vegetation, snow, etc.)

Planning and control modules are being adapted for rough terrain, with tighter coupling between perception and control to ensure stability in challenging conditions.

Dataset Integration and Training Progress

To support robust off-road perception, CoE members have combined six open datasets — including Rellis3D, Goose, ORFD, Yamaha-CMU, CaSSeD, and OFFSED — under a unified labeling scheme. A custom parsing and configuration framework has been developed to standardize classes and formats across sources.

Early training results are promising:

• Free space segmentation models have reached ~80% IoU, showing reliable detection of drivable areas.

Work continues on object detection and terrain classification, with initial tests already demonstrating generalization on unseen data.

While the dataset size (~22k images) remains smaller than typical on-road datasets, discussions are underway with potential industry contributors to expand the pool of off-road training data.

Simulation Environments

The Off-Road WG is extending Autoware’s simulation capabilities to cover both terrestrial and space-focused use cases:

• Mining environments: New vehicle and terrain models are being developed in CARLA 0.10, enabling realistic testing of heavy-duty off-road applications.
• Off-road racing: A scalable Isaac Sim RoboRacer model is being upgraded from 1/10th to 1/5th scale. A configurable racetrack generator has been built, allowing randomized parameters such as track width, banking, and corner angles.
• Mars environment: A digital twin has been created based on NASA Perseverance rover locations, with terrain models verified against rover images.
• Lunar environment: An environment based on a candidate Artemis landing site at the lunar South Pole has been developed, complete with detailed terrain and rock assets.

These environments provide realistic testbeds for validating perception, planning, and control pipelines in conditions ranging from industrial sites to planetary surfaces.

Collaboration Opportunities and Next Steps & Suggestions

The August meeting also highlighted upcoming opportunities for joint work and cross-university collaboration:

• Poznan University : Preparing a real off-road project with potential EU funding, including a 700 m² covered test facility with basalt sand for lunar rover analog studies.
• Virginia Polytechnic Institute and State University : Collecting small-scale airport data and opening access to off-road test sections. Alignment with Space ROS standards is in progress through NASA’s Picnic team.
• Standardized Platforms: Suggestions were raised for creating a common, affordable platform, similar to F1TENTH, to accelerate off-road autonomy research.
• Simulation & Physics: Interest in converging simulators and advancing physics models for granular media was shared by several members.
• Dataset Guidelines: Discussions began around tiered dataset requirements to ensure consistency across different projects.

The August CoE meeting marked a strong restart after the summer break, with new research results, simulation progress, and cross-university collaborations pointing toward impactful outcomes. From personalized autonomy and scalable vehicle dynamics to lunar and Martian testbeds, the CoE community continues to expand the boundaries of what Autoware can achieve in research and education.

We look forward to the September meeting and the next wave of updates from our university partners.

The post Autoware Centers of Excellence Steering Committee August 2025 Update appeared first on Autoware.

It’s been a while since our last newsletter and we’ve got a lot to catch up on! Over the past six months, the Autoware Foundation community has been busy advancing open-source autonomy, growing our member base, and pushing exciting new projects forward.

So, what can you expect from this special recap edition? Here’s a quick preview:

Major announcements shaping the future of open autonomous driving

New members joining our community and bringing fresh expertise

Highlights from events, workshops, and meetups worldwide

Updates on technology and demos

A look at how we’re expanding collaboration across industries and academia.

Whether you’re new to Autoware or a longtime supporter, this newsletter will give you a clear snapshot of where we are today and where we’re headed next.

Grab a coffee and dive in. There’s a lot to explore!

Major Announcements

Autoware Foundation 2.0 Has Arrived: A New Chapter Begins

Since its founding in 2018, the Autoware Foundation has grown into a vibrant global community, now spanning over 20 countries, supporting more than 30 vehicle types, and engaging with over 500 organizations worldwide.

With this growth comes a renewed commitment to governance and collaboration. To better serve the community and accelerate progress, the Autoware Foundation has refreshed its management structure under Autoware Foundation 2.0:

• Clear separation of monitoring and execution
• Smarter decision-making with broader expertise
• Closer teamwork across technical, strategic, and academic groups
• More business opportunities for members

Leadership updates:

Board of Directors:

• Yang Zhang (Chair), Christian John, Suraj Gajendra
• Founding Chair Shinpei Kato now serving as Fellow

New Execution Leadership Team:

• Daisuke Tanaka (President & Operations Lead)
• Ryohsuke Mitsudome (Technical Steering Committee Chair)
• Johannes Betz (Strategic Planning Committee Chair)
• Muhammad Zain Khawaja (Managing Director of Product)
• Steve Taylor (Managing Director of  Strategy)
• Samet Kütük (Head of Marketing / Community Advocate)

Thank you to everyone driving Autoware’s success.

Newest Members

Welcoming New Partners to the Autoware Community!

NIT Academy

With over 20 years of experience in engineering education and industry training, NIT Academy will collaborate on enhancing Autoware’s safety and reliability through upskilling contributors in automotive safety standards and certification.

scaledrive.ai

This AI startup focuses on software that helps robots understand and predict human behavior, aiming to boost Autoware’s prediction modules and support deployment in dynamic environments.

Weichai Power Co., Ltd

A leading commercial vehicle manufacturer in China, Weichai joins as our first OEM member, contributing to expanding Autoware’s capabilities in new operational domains, including off-road and mining applications.

Alternative Embedded Innovation AG

Based in Switzerland, AEI specializes in safety-critical embedded software for automotive and robotics applications, helping productize Autoware on top hardware platforms.

ASAM e.V.

Joining as an alliance partner, ASAM shares our vision of open standards and collaboration, supporting Autoware’s simulation and software-defined vehicle initiatives.

Inceptio Technology

A pioneer in autonomous trucking with over 200 million kilometers of commercial operation, Inceptio joins as a Premium Member to drive innovation and foster global partnerships.

Autoware Café

Fresh Conversations on Open-Source Autonomy

We’re excited to launch The Autoware Café, a brand-new series where we sit down with members of the Autoware community to share insights, stories, and the latest updates on all things Autoware.

In our first episode, we had the pleasure of chatting with Muhammad Zain Khawaja, Managing Director of Product at the Autoware Foundation. Zain took us through the exciting Privately Owned Vehicle initiative—a greenfield project delivering an AI-first, open-source autonomous driving stack designed for OEMs and Tier 1 suppliers. From L2+ to L4, powered by cutting-edge AI models like SceneSeg and EgoLanes, the conversation covered the product roadmap, upcoming features, and go-to-market plans.

If you missed it, catch the full podcast episode here: Watch the Autoware Café Episode 1

And the café is brewing again!

Join us next Wednesday, July 16 at 11 AM CET, for a thrilling new episode: “Putting an autonomous vehicle to the German driver’s license test!”

Our host, Samet Kütük, will be welcoming Ilir Tahiraj and Dominik Kulmer from the Technical University of Munich to tell us all about their daring experiment, live from the IAA MOBILITY show in this September. Previously, EDGAR wowed crowds at Oktoberfest – now the TUM team is aiming even higher by giving their favorite “driver” a real license (so it can tackle city errands!).

What to expect:

• Why TUM chose Autoware as their R&D software framework
• Insights on the current state of autonomous driving in Germany and the future ahead

Bring your questions, there’ll be plenty of time for Q&A!

Register for the LinkedIn live event on July 16th here.

We can’t wait to see you at the next Autoware Café!

Member & Community Spotlight

Showcasing Innovation and Collaboration Across the Autoware Ecosystem

EDGAR’s Munich Challenge is On!

FTM Institute of Automotive Technology TUM’s Autoware-powered EDGAR Wiesn Shuttle is back, this time aiming to pass the real German driver’s license test at IAA MOBILITY! The TUM team invites all autonomous vehicle developers to join this exciting challenge on Bavaria’s streets.

Want to see how your self-driving platform stacks up?

Enroll now and get ready to prove your vehicle’s capabilities: Join the EDGAR Challenge (Participation in the  challenge on September 9–14 requires all necessary permits and approvals from local authorities. )

Plus, the TUM team will share all driving data and open-source code after the event.

Watch the video here: EDGAR in Action – Video

Autoware Open AD Kit Now on Arm Learning Paths!

The Autoware Open AD Kit is now part of the Arm Learning Paths, offering developers a comprehensive guide to running demos on Arm Neoverse processors. This learning path covers SOAFEE architecture fundamentals, containerized workload deployment using Docker, and advanced configurations for cloud and on-premise environments.

Ideal for those familiar with Docker and looking to deepen their understanding of software-defined vehicle development on Arm platforms.

Start your learning journey here Arm Learning Paths

Autoware Open AD Kit Featured in ASAM Guide

We’re honored to be highlighted in the recently published ASAM Guide on Software Defined Vehicles (SDV). This guide recognizes Autoware Foundation as a key player alongside industry leaders like AUTOSAR, COVESA, and Eclipse Software Defined Vehicle.

Our collaboration with ASAM e.V. focuses on integrating their SOVD standard with our Open AD Kit initiative. This partnership aims to develop a demonstrator showcasing how open-source software and industry standards can work together to accelerate safe, interoperable autonomous driving.

Download the electronic copy here ASAM Guide PDF

Red Hat’s SOAFEE Blueprint Showcases Open AD Kit Integration

Red Hat’s AutoSD seed project recently showcased a blueprint integrating Autoware Foundation’s Open AD Kit with their automotive-grade OS layer, AutoSD. This integration supports Kubernetes-style deployment using Podman and Systemd, providing a container-first approach for in-vehicle systems.

Built on CentOS Automotive and Red Hat In-Vehicle OS, AutoSD delivers ISO 26262 ASIL-B safety certification, SELinux security, real-time workload support, and OTA lifecycle management — all critical for production-ready autonomous vehicles.

The blueprint also features Jumpstarter tooling for automated CI-friendly deployment on both real hardware and virtual platforms, simplifying testing and development workflows.

Watch the video and read the detailed blog here Red Hat Blueprint Video & Blog

Our community’s innovation and collaboration continue to push the boundaries of autonomous driving technology. Thank you to all contributors and partners driving this journey forward!

Want to see your update featured in the next Member & Community Spotlight?

Send your news, achievements, or project highlights to samet.kutuk@autoware.org — we’d love to share your story with the community!

Latest Tech Blogs

Exploring Zenoh with Autoware: Faster, Smarter, and More Secure Data Handling

Zenoh is gaining momentum in robotics, IoT, and automotive for its speed, efficiency, and security in data communications. As an alternative to traditional ROS middleware, Zenoh continues to evolve, with recent updates rolling out in its 1.0.0 (Firesong) release.

We’re excited to share a new blog post from ChenYing Kuo (CY) of ADLINK Technology and ZettaScale Technology—an active contributor in the Autoware community. In this post, CY walks through setting up a prototype Fleet Management System (FMS) to control multiple Autoware-powered vehicles in the CARLA simulator.

Zenoh acts as the communication backbone, connecting vehicles via AD APIs and efficiently publishing messages to the FMS server backend, enabling seamless orchestration over the internet.

Dive into the details and learn how Zenoh enhances Autoware’s multi-vehicle management here: Read the blog on Autoware Foundation Blog

Past Events

Highlights from the Road: Autoware Foundation Around the Globe

Reflecting on #CES2025: A Showcase of AI and Automotive Innovation

CES 2025 was a vibrant display of AI-powered tech shaping the future of software-defined vehicles (#SDV). Despite some major OEMs being absent, Japanese and Chinese automakers, tech giants like Microsoft and AWS, and many industry players filled the halls. The Autoware Foundation proudly sponsored the social event at the TIER IV booth, welcoming over 100 automotive professionals to experience live demos and connect with the community. We’re already looking forward to CES 2026!

Embedded World 2025

Though Autoware Foundation did not have a booth, many members like AMD, TIER IV, and Synopsys showcased their work on the show floor. Highlights included AMD’s automated valet parking demo powered by Autoware, TIER IV’s sensor fusion suite, and Synopsys’ Open AD Kit flavor. The event underscored the importance of openness and collaboration in automotive innovation.

Tech.AD 2025

This broad autonomous driving conference brought together a diverse group of industry experts. Topics ranged from radar technologies and automotive safety standards to cloud-native platforms. ASAM e.V. representatives played a key role, reinforcing the need for harmonized standards and collaborative development.

Meetup at the AWS Santa Clara Prototyping & Innovation Lab

Coinciding with NVIDIA GTC, this meetup featured presentations on the Autoware Open AD Kit demo, ASAM’s North American activities, and advancements in AI foundation models for the Privately Owned Vehicle stack. Supported by AWS, ASAM, Deepen AI, and TIER IV, it fostered technical discussions and community engagement.

COLUMBUSS Consortium Meeting

This consortium meeting highlighted the exclusive use of Autoware software for autonomous public transport in the Netherlands. Research institutions showcased driver training simulations and safety metrics, exploring shadow driver modes that blend human and autonomous control. The event also shed light on challenges and opportunities for off-road autonomy.

5th ICA Summit

Our Senior Tech Lead, Muhammad Zain Khawaja, moderated a panel on redefining vehicle software systems at the ICA Summit, joined by industry leaders from FORVIA, Porsche’s MHP, and HERE Technologies. This event was a key opportunity to showcase Autoware software and discuss the Privately Owned Vehicle stack with OEMs, Tier 1 suppliers, and startups.

IEEE MOST Conference

Autoware featured prominently with a tutorial and Open AD Kit demo led by Boyang Tian. Presentations included innovative uses of V2X technology to enable shared perception environments, alongside the unveiling of the Star Shuttle by The CAR Lab.

SOAFEE APAC Seminar

The seminar highlighted the pre-launch video of the Open AD Kit Blueprint (SOAFEE-branded version coming soon!) and featured Autoware in over half of the presentations and demos. This demonstrated clear traction for Autoware within the APAC region’s autonomous vehicle ecosystem.

ICRA 2025

The Roboracer Foundation hosted its 24th autonomous racing competition, complemented by the third Roboracer online SIM racing league, attracting 58 teams. Autoware community members, including newcomers from Duckietown Robotics, connected and collaborated throughout the event.

16th ITS European Congress

The COLUMBUSS project showcased their Autoware-powered autonomous bus, offering free-of-charge autonomy software tailored for public transport. The demonstration attracted considerable attention and underlined Autoware’s growing role in public transit innovation.

ADAS & Autonomous Vehicle Technology Expo

Although a pavilion booth was not present this year, Autoware Foundation members joined a strong community presence. Highlights included presentations on the Privately Owned Vehicle stack and Open AD Kit, alongside collaborations with partners like Deepen AI and ASAM e.V. The event reflected ongoing challenges in the German automotive sector and reinforced the need for open-source collaboration.

Electrified and Autonomous Systems Dive-In Day

Hosted by Switzerland Innovation Park Zurich, this full-day event featured Autoware vehicles and sensor-instrumented scaled models. Discussions covered regulatory frameworks and the future of electrified and autonomous systems, culminating in an informal barbecue fostering community connections.

Upcoming Events

Where to Find Autoware Foundation Next Months

July 29 Autoware Meetup @ AWS Prototyping & Innovation Lab, Santa Clara, USA

Aug 27 AVTE California, San Jose, USA

Sep 9 IAA Mobility, Munich, Germany

Sep 9 ROSCon Japan, 2025 Nagoya, Japan

Sep 10 ROSCon JP, Meetup Nagoya, Japan

Sep 25 GSVF 2025, Graz, Austria

Oct 27 ROSCon 2025, Singapore

Community Announcement Curation

Latest Updates from the Autoware Community

Download the Latest Open AD Kit Handout!

We’re excited to share the newest Open AD Kit Handout, featuring key updates revealed at #CES2025!

Highlights include:

• Integration of advanced perception containers, demonstrating cloud-to-edge software transition
• Red Hat’s AutoSD project booting on Open AD Kit within virtual prototyping environments with ASIL-B safety certification

Want to explore more?

Download the handout here: Open AD Kit Handout

Autoware Challenge 2024 Results Are In!

The Autoware Challenge 2024 wrapped up in January, sponsored by TIER IV. Congratulations to the winners:

Team CommonRoad — Effortless Rapid Prototyping for Motion Planning

Team AV Perception — Pedestrian Trajectory Prediction Models

Team GT-ARC, TU Berlin — Multi-sensor 3D Object Detection

Honorable Mention: Team FTM Mapping — Consistent Long-Term Point Cloud Mapping

Presentation recordings are now available!

Watch the full playlist here: Autoware Challenge 2024 Playlist

Autoware Foundation General Assembly 2024 Playlist Available!

Our annual General Assembly brought members and partners together to showcase diverse Autoware implementations and strategies.

Dive into the presentations and learn how Autoware is powering innovation worldwide.

Watch here: General Assembly 2024 Playlist

Meet Scene3D

Scene3D is an open-source neural network turning regular cameras into depth sensors, creating high-resolution depth maps with sharp object boundaries. It excels in tough scenarios like construction zones and adverse weather, adapting globally without extra training.

Why Scene3D matters:

• Full transparency and no licensing fees
• Real-time depth understanding prioritizing critical scene elements
• Reliable performance in all weather conditions

Learn more in this video narrated by Muhammad Zain Khawaja: Scene3D Video

Introducing SceneSeg

SceneSeg detects every critical object in raw camera feeds, from cars to unexpected hazards like loose tires, providing general-purpose vision intelligence that adapts worldwide.

Why SceneSeg is a game-changer:

• Open source freedom with community-driven improvements
• Learns what “important” looks like without custom detectors
• Handles rare, edge-case safety scenarios effectively

Tested across cities worldwide in varied conditions, SceneSeg is set to redefine autonomous vehicle perception.

Check out the narrated video by Muhammad Zain Khawaja here: SceneSeg Video

Thank you for catching up with all the exciting progress and community highlights from the Autoware Foundation. Our journey continues to grow stronger thanks to your passion, collaboration, and innovation.

Stay connected with us across all our channels to keep up with news, events, and exciting developments as they happen.

We look forward to sharing more progress and stories with you in the next newsletter. Until then, keep innovating and pushing the boundaries of what’s possible!

See you next time!

The post Autoware Foundation Newsletter: January to June Recap Edition! appeared first on Autoware.

Welcome back to the Autoware Centers of Excellence Updates. In November 2023, this is our second edition of the CoE newsletter.

Autoware Centers of Excellence are bustling with activity. Here are some of the highlights from last month.

The University of Macau joins the Autoware Centers of Excellence network!

We welcomed Prof. Cheng-Zong Xu from the University of Macau as a new addition to Autoware CoE directors. Dr. Xu’s research interests lie in parallel and distributed computing and cloud computing. He is a Chief Scientist of the Key Project on Smart City of MOST, China, and a Principal Investigator of the Key Project on Autonomous Driving of FDCT, Macau SAR.

Visit Dr. Xu’s webpage for more information.

Additionally, Dr. Xu introduced his work to the CoE members at the monthly CoE call, and it already sparked interest in collaboration with other academia members at the Autoware CoEs.

Arizona State University joins the Autoware Centers of Excellence network!

We welcomed Prof. Junfeng Zhao from Arizona State University as a new addition to the Autoware CoE directors. Dr. Zhao’s research interests particularly include connected and automated vehicles (CAV), CAV simulation and system integration, motion planning and controls, electrified propulsion system controls, and intelligent transportation systems.

Visit Dr. Zhao’s webpage for more information.

Another key point is that Dr. Zhao introduced the BELIV (Battery Electric & Intelligent Vehicle) lab he founded at Arizona State University. At his lab, Dr. Zhao and his team have an autonomous research vehicle, and they’ve already got to work on adapting the drive-by-wire systems for Autoware use.

Introduction to AutoDrive, a simulator for autonomous research and education

Chinmay Samak and Tanmay Samak from Clemson University presented their work on AutoDrive, a simulator for autonomous research and education. This simulator has been used for developing autonomous parking, behavioral cloning, and multi-agent planning.

At the Autoware Foundation LinkedIn feed, we mentioned their work before, but the tinkering twins continue to push the boundaries of simulation. Many academic publications and project competitions accredit their work, and they continue to improve their simulation platform.

Here are some of the must-reads regarding the AutoDrive platform:

1. Announcement of the AutoDrive platform

2. AutoDrive for F1Tenth simulator

And don’t forget to watch the AutoDrive pitch video!

Autoware has been invited to organize a track at the IEEE Serious Open Source Summit

An exciting event is coming up!

To be held on February 20-21, 2024, at the Computer History Museum in Mountain View, California, the IEEE Serious Open Source Summit brings together industry leaders, practitioners, and visionaries who are leveraging open-source technology to drive digital transformation in ClimateTech, Telecommunications, Manufacturing, Healthcare and Life Sciences, Finance, CivicTech, the Metaverse and other verticals.

Hosted by the IEEE Standards Association | IEEE SA, this event will showcase open source that forms the foundation of business operations, data science, AI, IT, sustainability, and IoT efforts and provide an opportunity to learn about the IEEE SA Open collaborative ecosystem designed to bridge the gap between open source communities and industry applications with a growing set of projects, standards, and initiatives.

Autoware Foundation was invited to organize a track at this wonderful event, and accordingly, we’re working on representing Autoware’s work with the motto “running a commercially successful business using open-source.” Undoubtedly, it will be interesting with invited speakers, presentations and a panel discussion.

For more information about the event, visit the event website.

Kicking off the Autonomous Logistics business case exploration work by Wharton School and the Autoware Foundation!

As mentioned earlier on our social media, The Autoware Foundation has been working with The Wharton School on several special projects (we will soon make a recap of those activities). One is the exploration of autonomous logistics business cases where the team will focus on autonomous material transportation for indoor/outdoor intralogistics in large production and manufacturing facilities.

Some of the deliverables will be

1. Customer discovery in the US
2. Market development and market-entry strategy
3. Pilot customer deployments in large industrial sites.

Additionally, this activity is built on top of the good work done at the Cargo Delivery ODD times, and it clearly shows interest and a path towards commercialization of solutions built on open-source technologies.

Before closing…

Autoware Centers of Excellence network is bustling with activity, by research, development, and deployment alike!

You don’t want to miss the third edition of the newsletter. Don’t forget to subscribe, like and share!

See you next time!

The post Autoware Centers of Excellence Updates November 2023 appeared first on Autoware.

Welcome to the third edition of the Autoware Technical Steering Committee Update newsletter!

Here’s a concise summary of what happened last month at the technical working groups of the Autoware Foundation.

Software Working Group

• In last month’s newsletter, we mentioned that the folks at Industrial Technology Research Institute (ITRI)(工業技術研究院, 工研院) have completed their Bus ODD demo shooting in Taiwan. As the work in editing and mixing continued withing the last month, we anticipate that we will be able to release the second Bus ODD demo video on Autoware Foundation’s YouTube channel in a few weeks. Stay tuned! Huei-Ru Tseng
• In Autoware, we used a CUDA version from Ubuntu 20.04 with a workaround because NVIDIA did not release the TensorRT for Arm until recently. Now that it’s released, we started testing the CUDA 12.3 version with x86-based and Arm-based systems and validated that the update works well. This update also reflected positively on the Autoware Docker image size. Also, Ansible scripts are updated successfully. M. Fatih CIRIT, Esteve Fernández, Oğuz Kağan Öztürk
• Built on top of last month’s activity, M. Fatih CIRIT released the Launch Unifier Tool to simplify the launch structure of the Autoware project. This will make understanding Autoware’s launch structure easier and streamline debugging. Make sure to check it out!

Localization & Mapping Working Group (as a sub-working group)

• The localization & mapping working group went through our roadmap spreadsheet and reorganized the working group’s goals towards giving it a finer granularity. At the Strategic Planning Committee, we will soon compose and reflect these feature goals in a release format. Work in progress; stay tuned!

Planning & Control Working Group (as a sub-working group)

• The planning & control working group continued working on running on hilly road testing (#5391, #5121) created by the community after their vehicle and simulation testing.
• Also, the working group is working towards improving Autoware’s planning behavior against obstacles (#3964), incorporating perception improvements for clustering and tracking-based detection of obstacles.
• Finally, the working group is studying various planning presets (#680) for different use cases, which will improve the ease of use and flexibility by switching launch modules.

Sensing & Perception Working Group (as a sub-working group)

• The sensing & perception working group is working on adding training documentation for the lidar centerpoint package. The documentation explains how to train custom CenterPoint models and convert them into applicable formats for deployment in Autoware. The added documentation will be released soon. PRs are in place. Kaan Çolak
• The working group also improves the image segmentation-based point cloud filter (#3974). Detection of some objects, such as growing plants and bushes on the roadside or discarding point clouds resulting from raindrops, exhaust gases or such non-trivial elements, is a struggle.

Operational Design Domain (ODD) Working Group

• Since the latest Technical Steering Committee meeting, a total of 554 scenario tests were performed on Autoware CI/CD, and 502 out of 554 tests were passed. The remaining slight number of failed cases are due to left turn cases, and they are being reviewed.

• A new scenario is added to the automated testing called “unprotected right with limited visibility.”
• Leo Drive contributed to port TIER IV’s basic scenarios on the virtual maps. There will be 374 new scenarios; however, due to the cloud resource limitation, new scenarios may be tested at a lower frequency than weekly testing. As of November 12, an initial 130 new scenarios were uploaded and tested.

Safety Pool joined the ODD working group meeting on November 20, and they plan to provide the comprehensive safety scenario database to Autoware Foundation. Also, they work closely with ASAM e.V.

Simulation Working Group

• OpenSCENARIO support for AWSIM is upcoming! The feature is moved to the public AWSIM repository and PR is waiting to be merged.
• The design of the verification procedure for the AWSIM and Autoware project continues. Scenario simulator and AWSIM are being tested on ITRI scenes since the ODD working group prepared ITRI LL2-based scenarios.

Racing Working Group

• Autoware Universe perception packages debugged and tested working (achieved real-time, 20-30 FPS multi-object detection and tracking with camera input) on F1Tenth’s Jetson Xavier NX. It will be released to the F1tenth branch shortly. F1Tenth Foundation
• The Autoware CoE team at the University of Pennsylvania is exploring various simulation environments for the AV4EV go-kart platform. The team is collaborating with AlexanderRex, who built the AV Go-Kart simulation, which is based on Unreal Engine and the CARLA simulator

Open AD Kit Working Group

• Software portability continues to be a major drive at the Open AD Kit working group. WG participants from semiconductor companies are using Autoware to prepare demos for the upcoming #CES2024 show.
• Ad-hoc meetings are continuing to assess the performance of the very granular containerization of Autoware. Interesting preliminary performance comparisons are observed, and discussion and further experimentation will continue. Release plans envisioned but pending announcement.
• DevOps Dojos effort is continued -> DevOps Dojo dashboard
• The ROS Node update effort is continued (a call for volunteers is in place) -> ROS Node update status spreadsheet. We also created a liaison with the ROS community where Kasper Mecklenburg will introduce ROS-related efforts at the Open AD Kit WG to pull in more attraction and resources from the ROS community.
• CES demo preparations are ongoing successfully. Open AD Kit demos will be exhibited in at least three spots at the upcoming CES show.

Reference Design Working Group

• Discussions on the organization of reference design documents were continued.
• The group reviewed two reference design drafts from F1Tenth and AV4EV Go-Kart
• Created a draft schematic overview for reference design document. Here’s a draft of the meta-metric matrix.

API Working Group

• The API working group published a sample API definition for review.
• The working group created a demo of integration with a DANLAW OBU and Autoware project, showcasing a use case where the autonomous vehicle changes lanes when an emergency vehicle starts broadcasting its position and lane closure notification.

• A use case on virtual traffic light interface using V2X was also presented
• Since the proposed API definition is accepted, all the relevant APIs will be defined in the coming WG meetings.

Don’t forget to come back for the next TSC Update newsletter!

And please subscribe to this newsletter and other Autoware Foundation LinkedIn newsletters!

We also started the Centers of Excellence newsletter as well. Check it out here.

There is just so much going on at the Autoware Foundation. Make sure to like/repost our LinkedIn content and visit our website at autoware.org

The post Autoware Technical Steering Committee Update – November 2023 appeared first on Autoware.

Welcome to the first edition of the Autoware Centers of Excellence newsletter!

As we created the Autoware Foundation Technical Steering Committee newsletter that goes out every month, we will be publishing the Autoware Centers of Excellence newsletter every month as well.

Autoware Centers of Excellence initiative grew to 22 universities and research institutions spanning three continents. We have a lot of exciting activities as a part of the Autoware CoE network that could help in understanding Autoware activities from a research perspective.

Autoware for OpenEV Platforms

The Autoware Centers of Excellence, headquartered at the University of Pennsylvania, has been working on the development of various vehicle platforms to instrument them with autonomous driving capabilities. These vehicles span from 1/10th scale F1Tenth vehicles to 1/2 scale research platforms to full-size EV logistics vehicles. The main idea is to homogenize the development of these vehicles by using virtually the same open-source software and system architecture.

Autoware Centers of Excellence continues to expand

Autoware Centers of Excellence is now at 22 universities and research institutions, spanning three continents. All the CoE member universities and research institutions use Autoware in their development platforms, which creates a great synergy between academia to work on a joint platform and great research output for Autoware.

Autoware CoE organizes exciting racing and community events across the globe!

Autonomous racing is a great instrument for university students and professors to get involved with real hands-on autonomy development. This brings much exposure to the Autoware project, and Autoware CoE is skillfully coordinating these efforts. The recent engagements are mostly based on 1/10 scale F1Tenth vehicles due to Autoware Foundation’s F1Tenth Foundation affiliation. Still, Autoware is also utilized for other racing vehicles (such as the Indy Autonomous Challenge vehicles), and Autoware CoE’s involvement in autonomous racing will only get more broadscale in the near future.

F1Tenth Competition at IROS 2023

The F1Tenth Competition at IROS 2023 was held on October 2-4. Autoware teams running the Autoware stack on the F1Tenth vehicle came in second and third place in the competition, while the undergraduate team from the University of Quebec finished the competition in first place. This was a very fun and exciting event; it has been the center of attention for the entire conference.

F1Tenth Competition in South Korea

The F1Tenth competition in South Korea was a huge event, with 31 teams included (29 of them raced) in the races. The teams joined the competition with the same reference vehicle design shared by the F1Tenth organization (as a part of UPenn’s Safety21 project), and this evolved to be a national program in South Korea. 20+ universities funded their teams, and they will continue to do so in the coming two years as well.

Po-Jen Wang joined the ROS By-the-Bay meetup in Bay Area

Po-Jen Wang, a software developer at the Autoware Foundation, joined the ROS By-the-Bay, the exclusive Bay area meetup for ROS enthusiasts.

Po-Jen’s talk was titled “AV4EV – Open-source AV software for open-standard EV platforms”. He reviewed Autoware CoE activities at Intrinsic’s office with the ROS community.

The MAD (Multi-Agent Dynamic) Games

The Autoware CoE team contributed and participated in the 1st Multi-Agent Dynamic Games workshop during the IROS 2023 conference. The main objective of the MAD Games workshop is to come up with various vehicle driving strategies to realize safety through agility. These are very adversarial games, so how to navigate intersections, roundabouts and risky situations were the main topics explored. This is a great exercise to study the state-of-art for autonomous driving technologies/methodologies.

---

The Yamaha EV platform is ready to roll!

Autoware CoE HQ at the University of Pennsylvania has received a Yamaha EV platform (in collaboration with eve autonomy) to develop and productize the Autoware stack for indoor-outdoor intralogistics applications. The Eve Autonomy vehicle was demonstrated (as well as the 1/10th scale F1Tenth vehicle and 1/2 scale autonomous go-kart vehicle) during the Fall 2023 GRASP Industry Day (UPenn’s annual Robotics Industry Day).

---

That’s it for the first Autoware Centers of Excellence newsletter. Don’t forget to come back next month with more updates.

The post Autoware Centers of Excellence Updates – October 2023 appeared first on Autoware.

Welcome to the second edition of the TSC Update newsletter!

Here’s a concise summary of what happened last month at the technical working groups of the Autoware Foundation.

Software Working Group

• Folks at the Industrial Technology Research Institute (ITRI)(工業技術研究院, 工研院) have completed their Bus ODD demo shooting in Taiwan, and we will be releasing the second and final edition of the Bus ODD work very soon.
• A great effort in building various GUIs to make the Autoware project more convenient to use has been started by Khalil Selyan – See here -> The Build GUI | The Launch GUI

• AWF Software Architect M. Fatih CIRIT prepared a beautiful Launch Unifier Tool to simplify the launch structure of the Autoware Project. Definitely worth taking a look, don’t miss it

Localization & Mapping Working Group (as a sub-working group)

• This month, the work continued on Georeferenced Point Cloud Map Generation, Dynamic Object Removal and Localization Error Metrics.

Localization & Mapping Working Group (as a sub-working group)

• This month, the work continued on Georeferenced Point Cloud Map Generation, Dynamic Object Removal and Localization Error Metrics.

Planning & Control Working Group (as a sub-working group)

• Some improvements are being discussed to the longitudinal controller for hilly road tests.
• Requests are made to support overhanging obstacles during the filtering process.
• Addressing some of the issues that arose from AWSIM tests continued.

Sensing & Perception Working Group (as a sub-working group)

• We started to prepare training codes for perception nodes
• Added RoboSense LiDAR support in universal LiDAR driver
• Implemented dust detection process for LiDAR point clouds using morphological processing.

Operational Design Domain (ODD) Working Group

• Since the latest Technical Steering Committee meeting, a total of 552 scenario tests were performed on Autoware CI/CD, and 366 out of 552 tests were passed. A startup script issue caused a total failure (0/138 succeeded) of the weekly test, and it was quickly fixed. Other issues were investigated, which was a side effect of the new planning module.

• The use case list has been brushed up, and some typical scenarios have been prototyped.
• For the pullover scenarios, the Shalun (Taiwan) map was chosen. The unconnected rightmost lanelets were assigned with the subtype of “road_shoulder” instead of normal “road”.

• Leo Drive proposed to provide the scenarios on a virtual map to enrich the test scenario database. A new “catalog (set of scenarios)” in the CI/CD pipeline is being prepared to accept the new scenarios.

Simulation Working Group

Design of verification procedure for AWSIM and Autoware project:

• The Wiki page is created with a description of the verification procedure
• OpenScenario library to be prepared (ODD WG scenarios  can be reused, possibly more to be created)
• Scenario Simulator (SS2) and AWSIM integration will be used to execute scenarios (the feature is ready and to be merged into public AWSIM soon

Racing Working Group

• Two teams (Autoware Aces & Autoware Rocket) used the Autoware project (Po-Jen Wang and Tejas Agarwal) during the 14th F1Tenth Autonomous Grand Prix at IROS 2023 in Detroit. They were both one-man teams, running their own F1Tenth vehicles. Po-Jen used advanced gap follow (no map), and Tejas used pure pursuit on an optimized trajectory (using a map), and standard gap follow. Po-Jen finished in second place, and Tejas finished in third place.

Autoware integration on F1Tenth continues:

• Currently testing the perception packages on Jetson: tensorrt_yolox and bytetrack
• tensorrt_yolox works well with real-time fps
• Currently resolving tracking issues with bytetrack

Open AD Kit Working Group

• Continuing to experience new engagement opportunities and input from prospective members. Colleagues from Red Hat started to join the working group.
• We started working on the CES demo using the Open AD Kit, which will be built on top of the demo we prepared for the Autonomous Vehicle Technology Expo in Stuttgart.
• Great discussions are ongoing for Reducing Autoware container size and optimizing developer and production environments. Check out the in-depth discussion and the progress.
• DevOps Dojos effort is continued -> DevOps Dojo dashboard
• The ROS Node update effort is continued (a call for volunteers is in place) -> ROS Node update status spreadsheet
• The open AD Kit demo at IROS 2023 in Detroit was completed successfully. The PIX Moving chassis was instrumented with the Ampere Altra Developer Platform (AADP) from ADLINK Technology and the CAN adapter from Kvaser AB. The chassis was running on the street autonomously, and a video was captured to be shown alongside the chassis in a static demonstration at the IROS show floor and it was received very well.

• The content created around this demo was among the most liked/commented/reposted content published on Autoware Foundation’s LinkedIn channel.

Reference Design Working Group

• The hardware reference design document was finalized.
• A reference design template was proposed, for example designs for various projects and platforms.

API Working Group

• The V2X interface design proposal was discussed at the working group meeting
• Also, the working group started to populate the Autoware Documentation Design AD API pages.

See you at the next TSC Update in December!

Don’t forget to subscribe to this newsletter and other Autoware Foundation LinkedIn newsletters!

We’re soon starting the Centers of Excellence newsletter as well.

There is just so much going on at the Autoware Foundation. Make sure to like/repost our LinkedIn content and visit our website at autoware.org

The post Autoware Technical Steering Committee – October 2023 appeared first on Autoware.

As the new year has now officially begun, it’s time to reflect on the past and look forward to the future.

Autoware Foundation starts the new year with a new Board of Directors!

The co-founders of Autoware, Shinpei Kato (TIER IV) and Yang Zhang (AutoCore.ai), remain on the board of the Autoware Foundation. They have been joined by Samet Kütük from Leo Drive (who also leads the Marketing/Go-To-Market workgroup), Rahul Mangharam from the University of Pennsylvania (who heads the Centers of Excellence initiative), and Larry Dimitrievski from Danlaw, all of whom have joined the Autoware Board of Directors to help bring the project to new levels of success.

In 2022, the Autoware Foundation made significant progress in efforts towards advancing the development and deployment of autonomous vehicles. Here are some of the notable achievements:

• Autoware moved back to GitHub, and the Autoware repository was restructured by merging the Autoware-AI repository.
• The Autoware Foundation introduced the Autoware Core and Universe paradigm and developed the Bus ODD project using the Autoware Universe.
• ITRI and Leo Drive conducted tests of the Bus ODD project using real vehicles. The Autoware Foundation is working on releasing demo videos showcasing these tests soon.
• Mapping Workgroup was created, and many activities around mapping and localization are being undertaken, such as GNSS/IMU fusion, dynamic map loading, sensor calibration tools, camera-based localization and highway mapping/localization.
• The Autoware Foundation has become a member of the SOAFEE initiative, while Open AD Kit 2.0 was developed and showcased at prestigious events such as MIH Demo Day and Arm DevSummit.
• AWSIM was released as open-source by TIER IV to be the default simulator for Autoware. Simulation Workgroup also works and integrates Autoware with other simulators such as MORAI Sim and Carla.

Some GitHub statistics:

• 1,653 Pull Requests merged
• 297 issues closed
• 133 unique contributors

To match these technical achievements, the Autoware Foundation made great strides in

• Revitalizing the foundation’s marketing outreach by working on a new Autoware website, blogging on LinkedIn and Autoware Blog, joining and organizing more than twenty events in three continents
• Expanding the Autoware Centers of Excellence initiative to 11 universities on three continents
• Strengthening its position in the Software-Defined Vehicle development paradigm through alliances
• Committing to fostering growth and success within the Autoware community with sincere dedication.

Thank you for supporting and loving Autoware!

The post Happy New Year from the Autoware Foundation! appeared first on Autoware.