By Doug Newcomb, AVCC Outreach Ambassador

From new technical reports and valuable industry collaborations to significant staff addition and a successful inaugural annual event, 2023 was a banner year for AVCC.

As 2023 draws to a close, we wanted to look back on a momentous year of growth and accomplishments at AVCC, a global consortium of automotive and technology industry leaders cooperating on intelligent software-defined and automated vehicle technology. And highlight the organization’s exciting achievements, events and milestones over the past 12 months. Plus, stay tuned because AVCC has much more in store for 2024!

Release of a New Technical Report on Models and Datasets

In January AVCC announced that its latest technical report – TR-004 Models and Datasets for Benchmarking Deep Neural Networks (DNN) for Automated and Assisted Driving Systems – was available to the public for free download. AVCC saw a need for the report since industry alignment on how to fairly compare DNN performance on compute platforms for automated and assisted driving systems was nonexistent.

Having a common view on how to benchmark machine learning (ML) in the automotive industry benefits the entire ecosystem. For the automotive and ML communities, this allows optimized use of models and datasets that are publicly available and effectively addresses the key usage examples of DNNs in real-world automated and assisted driving systems applications.

To download the report, visit https://www.avcc.org/tr004/

AVCC and MLCommons Join Forces to Develop an Automotive Industry Standard Machine Learning Benchmark Suite

In May, AVCC and MLCommons®, an open and global engineering consortium dedicated to making machine learning (ML) better for everyone, announced the industry’s first Automotive Benchmark to take specifications and benchmarks and move them through to open-source software and certification.

The partnership was formed to develop an industry-standard Automotive Benchmark Suite for use by OEMs and automotive suppliers using AI/ML Deep Neural Network (DNN) technology. The Suite will build upon the AVCC AI/ML Benchmark Technical Reports and the MLPerf™ benchmark suites developed by MLCommons. This common set of benchmarks will help guide the industry’s collective engineering around future platforms, accelerating the development of new capabilities.

The technical work began with initial participation in the benchmark from Arm (co-chair), Bosch, cTuning Foundation, KPIT, NVIDIA, Red Hat, Qualcomm, Inc. (co-chair), Samsung Electronics, and other industry leaders. To join the Standard Automotive Benchmark Suite initiative, visit: https://mlcommons.org/en/groups/inference-automotive/.

Automotive Benchmark Task Force – Call for Participation

Following the announcement of AVCC/MLCommons partnership, the organizations issued an open call for participation to industry stakeholders to join and be at the forefront of automotive vehicle technology benchmarking, with the goal of defining and developing an industry-standard ML benchmark suite to be used in automotive requests for information (RFIs) and requests for quotation (RFQs). AVCC has a proven track record of publishing Technical Reports on how to measure ML performance for automotive. Participants for the first time can leverage AVCC’s reputation for publishing Technical Reports on how to measure ML performance for automotive and MLCommons’ track record of developing well adopted industry standard ML benchmark suites to focus on defining the right AI/ML performance targets and specifications to ensure fairness, accuracy and broad industry adoption

AVCC Adds Car Tech and Event Expert Doug Newcomb to its Team as Outreach Ambassador

In June AVCC announced Doug Newcomb joined the organization as its Outreach Ambassador, bringing a wealth of knowledge and experience that reinforces AVCC commitment to advancing the development of automated and assisted driving technologies. As an influential thought leader and experienced journalist, Newcomb has been at the forefront of the automotive industry for over three decades. His deep understanding of emerging technologies and their impact on the transportation sector has made him a sought-after authority on autonomous vehicles, electric mobility, and connected transportation systems.

Some of Newcomb’s responsibilities include starting conversations and building relationships with different stakeholders in the automotive industry and connecting them to the Outreach Committee. Newcomb’s appointment strengthened AVCC’s efforts in shaping the future of autonomous vehicles through technological consensus around non-differentiable IP blocks.

AVCC Welcomes Prominent Mobility Expert Selika Josiah Talbott as Director of Special Projects

In June AVCC also announced the addition of Selika Josiah Talbott – a leading industry voice for responsible investment, engagement, and preparation for new mobility, including autonomous vehicles – as Director of Special Projects for the organization. With extensive experience in the transportation industry, Talbott brings invaluable insights and guidance to the organization. Having worked at the forefront of the industry for many years, she has a deep understanding of the complex challenges and opportunities in the development of autonomous vehicles. Her strong technical background, coupled with her strategic vision, legislative experience, and strong voice in the industry, has helped propel AVCC toward achieving its goal of becoming a key market enabler for intelligent-vehicle software-defined technology.

Among Talbott’s responsibilities is representing AVCC at selected tradeshows and conferences to establish the organization as the industry reference for AV software-defined automotive technologies. In addition to her new role as the Director of Special Projects for AVCC, Talbott is the Founder and CEO of Autonomous Vehicle Consulting.

AVCC Announces Strategy Update as Industry Collaborates to Accelerate the Deployment of Autonomous Vehicles

In July AVCC announced a broad strategy update designed to accelerate the development and deployment of automated and assisted driving technologies. Since its founding, AVCC has focused its efforts on successfully delivering technical reports for system architectures and requirements for automated and AV computing, including cybersecurity, functional safety, and building-block interconnects. Moving forward, AVCC has broadened its focus to include intelligent-vehicle software-defined technology and to extend its reach beyond the role of a traditional technical consortium. An example of this broader strategy is AVCC’s announcement of its partnership with MLCommons to develop the first automotive machine-learning benchmark suite.

AVCC’s new multi-prong strategy is designed to leverage industry collaboration and foster cooperation even between traditional competitors. It focuses its efforts on developing non-differentiable solutions that aggregate contributions from all its members and deliver solutions that benefit the entire community. This strategy enables its members to focus on differentiable IP building blocks while benefiting from economies of scale, reusability, and shorter time-to-market.

AVCC Joins SOAFEE, Expands Focus on Software-Defined Vehicles

Also in July AVCC announced it joined the Scalable Open Architecture for Embedded Edge (SOAFEE) initiative, an industry-led Special Interest Group (SIG) aimed at bringing open, standards-based architecture for the software-defined vehicle and modern software development techniques into the automotive industry. Founded by Arm and other industry leaders across the automotive supply chain, SOAFEE creates an opportunity for the traditional automotive industry and the software development community to come together and share their expertise, technologies, and products as the future of the software-defined vehicle continues to evolve.

This model of collaboration aligns with AVCC’s core belief that cooperation is the only path to AV success. As a new member of the SOAFEE SIG, AVCC will contribute to the goal of leading and defining the cloud-native development paradigm required for a new era of efficient edge workloads. The collaboration brings together expertise from diverse domains and advances their complementary areas of work.

The Future of Automated and Autonomous Driving Technologies – AVCC2023 Conference in Cambridge, UK

In August, AVCC announced its first-annual conference, AVCC2023: Driving the Future of Software-Defined Vehicles, which took place September 26 to 28 at Arm headquarters in Cambridge, UK. The conference featured a range of engaging presentations, panel discussions, workshops, social events and networking opportunities showcasing the latest advancements, innovations, and insights in automated and autonomous driving technologies.

The event program was carefully curated to provide an unrivaled opportunity to learn firsthand the groundbreaking research, collaboration and concepts shaping the future of automated and autonomous transportation. The event also featured an in-person meeting of the joint AVCC/MLCommons Automotive Benchmark Task Force.

AVCC Announces the Release of New Technical Report to Help Simplify and Accelerate the Development of Software-Defined Vehicles

In September AVCC announced the publication of its latest technical report, TR-005: Data-Oriented Communication Architecture for Automated and Assisted Driving Systems. The member-written multi-company technical report provides the auto industry with recommendations on common communication data architecture standards and best practices that enable an ecosystem of independently developed, reusable software components to streamline software development, reduce software system integration risks during their lifecycle cut costs and shorten timelines.

It also provides the automotive ecosystem with information they need to eliminate one of the most pressing challenges in building software-defined vehicles: How to manage increasingly complex data flows across evolving vehicle software systems while keeping the cost of software system development and integration to a minimum. By utilizing the recommendations in this technical report, developers and integrators can concentrate on other aspects of the vehicle software.

For more information or to download a complimentary copy of AVCC’s TR-005: Common Communication Data Architecture, please visit AVCC’s website at https://avcc.org/tr005/.

Arm Automotive VP Suraj Gajendra and Automated Vehicle Pioneer Seval Oz Deliver Keynotes at AVCC2023

In September AVCC announced Suraj Gajendra, vice president of Automotive Products, Software and Solutions at Arm, and Seval Oz, Founder of Oz Advisors and a Board Member for MicroVision and Pioneer Electronics, would deliver keynotes at the inaugural AVCC2023 in Cambridge, UK on September 26 to 28.

Gajendra’s keynote, titled ‘Four Wheels and a Billion Lines of Code’, covered software-defined vehicle trends, the role of Arm in SDV and how it aligns with the work the AVCC is doing. Oz’s keynote session was titled ‘Learning from the Past to Making AVs a Reality in the Present: No Shortcuts, No Dead-ends.’

Both keynotes are available for viewing on the AVCC YouTube channel.

AVCC2023: Driving the Future of Software-Defined Vehicles – Event Recap

Following its hugely successful first-annual conference, AVCC2023: Driving the Future of Software-Defined Vehicles, that took place September 26 to 28 at Arm headquarters in Cambridge, UK, AVCC posted a recap in photos and videos. Videos from the event – including keynotes, presentations, panel discussions, and more showcasing the latest advancements, innovations, and insights in automated and autonomous driving technologies – are available on the AVCC YouTube channel.

AVCC and MLCommons Gather for Signing Ceremony at AVCC2023

Wrapping up an exciting year, MLCommons and AVCC conducted a signing ceremony while together in person at AVCC2023 of their legal collaboration agreement, formalizing and consolidating the industry’s first Automotive Benchmark joint initiative. The suite builds upon the AVCC ML benchmark Technical Reports and the existing MLPerf™ benchmark framework developed by MLCommons, a common set of benchmarks that help guide the industry’s collective engineering around future platforms, accelerating the development of new capabilities, while the joint task force provides the industry with certified guidance to make informed decisions throughout the development process.

Looking Ahead to 2024!

2023 was an exciting year for AVCC and autonomous and automated vehicle technology. As we enter 2024, we have lots to look forward to, including a presence at CES, technical papers in the pipeline for Q1 publication, our annual member meeting in May and the second-annual AVCC2024 in September. But we’ll preview all of this and more in another blog post.

In the meantime, happy holidays from AVCC and all the best in the new year. We look forward to working with industry stakeholders in 2024 and beyond to assist in bringing key player together to develop consensus to scale innovation and quickly lower costs and help autonomous vehicle technology reach its potential to help reduce accidents, save lives, reduce congestion, and save fuel.