Recap of AWS re:Invent 2023 for the Automotive Industry

As we reflect on the impactful events of Amazon Web Services (AWS) re:Invent 2023, it’s clear that this year’s conference, which took place in Las Vegas, Nevada, was a nexus of innovation and forward-thinking, especially for the automotive and manufacturing sector. This recap is dedicated to distilling the wealth of information and announcements made during the event, specifically curating the content that stands to drive the automotive industry forward. From the AWS Automotive and Manufacturing Industry Business Unit (IBU) team’s insightful breakout sessions to the practical applications demonstrated in workshops and chalk talks, the resources provided at re:Invent 2023 are set to address the evolving challenges of the automotive field. As the industry accelerates towards a future of mobility which is safe, sustainable, and customer-centric, this summary will provide you with insights to navigate the transformation, ensuring a superior mobility experience for customers in the journey ahead.

At AWS, we are committed to accelerating the delivery of sustainable and safe mobility, products, services, and solutions to our customers through our unique collaboration and innovation and helping customers solve their transformational challenges. The Automotive and Manufacturing IBU within AWS works with several major OEM’s, Tier1’s, customers and partners and focuses on eight strategic workloads. This blog post focuses on automotive and manufacturing relevant announcements as presented during re:Invent distributed across eight strategic workloads. In each of the sections, we provide links to the re:Invent session recordings where they have been made available.

Software Defined Vehicles (SDV)

The domain of Software-Defined Vehicles (SDV) encompasses use cases that revolve around restructuring vehicle software to better access, understand, use, and update services essential for emerging mobility solutions. In the AUT102 breakout session, which gives an overview of the AWS vision for next generation cloud-first intelligent code pipelines for SDV, learn about how Traton and Volvo are accelerating time to market, while reducing development costs by creating a vehicle-to-cloud continuum, partitioning vehicle functions across different contexts and utilizing the BlackBerry QNX AMI for Amazon Elastic Compute Cloud (Amazon EC2) instances within development workflows. These innovations allow teams to identify bugs sooner, improve code quality, and reduce dependency on Hardware-in-the-Loop (HIL) systems. This is the testimony to shift left approach in the industry.

In the Chalk Talk session AUT203, accelerate and scale using cloud-native tools and virtualized targets introduced the latest cloud-native innovations for SDV architectures to help you build and test code in the cloud, before deploying it to the vehicle. The focus point of the discussion was the virtual engineering workbench, virtual electronic control units, and CI/CD pipelines and deep dive into use cases that demonstrate how generative artificial intelligence (generative AI) using Amazon CodeWhisperer can be leveraged to enhance productivity. Original Equipment Manufacturers (OEMs) and suppliers at various tiers discussed their current status of SDV adoption and outlined their forthcoming initiatives during the chalk talk.

The AUT301 workshop, “Automotive software development with Virtual Engineering Workbench (VEW),” was attended by over a hundred developers and architects. This workshop demonstrated how to build and test automotive-grade vehicle functions on AWS. In this builders’ workshop, participants created pre-configured AUTOSAR & QNX runtime environments and published them in the AWS Service Catalog. From the VEW self-service portal, users selected and logged in to their pre-configured AUTOSAR & QNX environments to develop a demo vehicle function application. Participants then integrated and executed vehicle applications on virtualized targets (virtual ECUs), running on Amazon EC2 instances, to verify and test functionality. Additionally, the workshop provided the end-to-end vision of the VEW, discussing how it can be expanded and tailored to meet the specific requirements and workflows of OEMs.

The Lightning Talk AUT103, “Accelerate automotive cockpit development with Panasonic SkipGen on AWS” explored how Panasonic SkipGen on AWS Graviton revolutionizes development for automotive cockpit domain controllers. This enables SkipGen’s ability to empower the industry by advancing development efficiency, enabling complete modern cockpit software development on AWS, compute offloading, and testing at scale in AWS.

AWS showed a demo about how the cloud-native Software Developer Workbench is scaling, and accelerating vehicle software development in the SDV era. In automotive experience at industry pavilion, AWS showcased eight demos, including an exhibit on the BMW i7 vehicle with information about its development on AWS.

Autonomous Mobility

In autonomous mobility area, AWS leverages Amazon’s years of experience with autonomous systems, robotics, and machine learning to accelerate autonomous vehicle development.

Breakout session AUT202 on highly automated driving with BMW and Qualcomm delivered learnings about how BMW is collaborating with AWS and Qualcomm to develop its next- generation highly automated driving development platform in the cloud.

Breakout session AUT206 on how Torc Robotics created a digital testing platform on AWS to run millions of miles in simulation, maximizing the test coverage of level 4 autonomy features, explores how they used AWS managed services to set up the platform. Speakers explored how Amazon Managed Workflows for Apache Airflow (Amazon MWAA), Amazon Elastic Kubernetes Service (Amazon EKS) and Amazon Simple Storage Service (Amazon S3) were used to set up the platform, technical challenges faced, lessons learned, and resulting benefits of the implemented solution.

PRO301 was a breakout session about next-gen trucking: How Iveco uses AWS to harness generative AI Autonomous driving to enable new heights of safety and efficiency. To realize this vision, Iveco is leveraging generative AI techniques built on AWS to redefine the relationship between drivers and vehicles. In collaboration with AWS Professional Services, Iveco is training customized generative models on privacy-compliant fleet data to elevate overall driver experience.

In the Chalk Talk session AUT302, the focus was on generative AI and Natural Language Processing (NLP)-based scene search in automated driving development. The discussion delved into the most significant challenges encountered in developing features for automated driving. It provided insights on how to search through petabytes of data to identify the relevant scenes for training and testing using generative AI. In this Chalk Talk, participants learned how generative AI helped accelerate scene selection, identifying the rare events and semantically similar scenes that are used in downstream tasks, such as ML training, testing, and validation.

Builder session AUT303 used generative AI to add objects within scenarios for model training in the ADDF solution. Automakers typically collect hundreds of petabytes of drive data from vehicle test fleets for autonomous and highly automated driving feature development, but sometimes do not log the exact scenarios needed by ML engineers to train models against corner cases. This builder session demonstrated how to add images and objects, such as stop signs, into existing scenes using generative AI.

At the demo booth, AWS showcased various tools available to support customers as they develop highly-automated driving features in the cloud. The demo booth showed AWS services used in support of data ingest, data pre-processing, scene generation, scene search, and large-scale re-simulation.

Connected Mobility

In the Connected Mobility area, AWS customers are unlocking the power of data to build intelligent, personalized features and revenue-generating mobility services. The AWS IoT FleetWise team announced its support for vehicle vision system data collection, enabling customers to collect metadata, object list and detection data, and images or videos from cameras, lidar, radar, and other vision sub-systems.

One particularly well-received breakout session, ALX201, focused on transforming the in-vehicle voice experience from concept to reality. Participants learned how BMW is building a unique, next-generation AI voice assistant that takes the in-vehicle voice experience to new heights. Regardless of internet connectivity, AWS enhances the BMW cloud environment by deploying an embedded neural text-to-speech SDK, that offers customers a highly natural, uninterrupted voice experience.

Another breakout session, IOT 204, delved into innovation and modernization of connected vehicle platforms with AWS IoT Connected Vehicle (CV) platforms. This session highlighted the potential for innovative applications such as AI-generated warranty and repair notifications, live video streaming and playback, EV battery health monitoring, and more. American Honda Motor Company shared their experience migrating to AWS IoT Core and outlined their plans for future innovations.

In the ANT 317 session, building real-time analytics from electric vehicles, Rivian shared connected mobility use cases on AWS. The Rivian vehicle data platform serves as a foundational service supporting various domains including digital commerce, insurance, advanced driver assistance systems, vehicle reliability, smart diagnostics, charging and vehicle servicing.

Breakout session IOT 309, innovate your applications using AWS IoT Core with MQTT 5 gave an overview of MQTT 5, and then explored connected vehicle use cases. Participants discovered how to communicate over potentially intermittent network connections using the publish and subscribe message features of MQTT. A live demonstration showcased how to scale fleet and applications using shared subscriptions.

In demo area, AWS team showcased the AWS Connected Mobility Solution 2.0 (CMS), illustrating how it facilitates the development, deployment, and management of connected mobility infrastructure at scale.

Digital Customer Engagement

In the domain of Digital Customer Engagement (DCE), AWS assists customers to increase customer engagement through personalized marketing content, immersive digital experiences, and real-time data analysis. This encompasses various aspects, including advertising, financing, after-sales support, and the repurchase experience, spanning the entire ownership lifecycle and customer journey.

In the AIM 206 breakout session, titled “Realizing value and business outcomes with Generative AI”, speakers presented convergence of AI and the human mind, coupled with technological advancements that are driving exponential innovation. In this session, participants learned how Ferrari is exploring generative AI with DXC Technology and AWS, the key impediments to generative AI adoption, areas of focus for mainstream adoption, and how to build value stream mapping for both internal and external consumers.

The Chalk Talk session AUT 204, titled “Driving into the future with Generative AI” discussed how Amazon Bedrock and Amazon CodeWhisperer are being adopted to enhance the digital experience throughout the customer journey, from pre-sales to post-sales, as presented by experts from the AWS Generative AI Innovation Center.

In demo area, AWS showed how generative AI is powering the Digital Customer Experience from call centers to predictive maintenance.

Manufacturing

In the Manufacturing sector, AWS is optimizing manufacturing operations and overall equipment effectiveness by capturing, analyzing, and visualizing data from the shop floor. The AIM 216 breakout session, “Predictive maintenance at scale”, shared Koch Ag & Energy Solutions (KAES) journey with Amazon Monitron. Unexpected equipment failure is costly for industrial facilities, while scheduling maintenance too frequently wastes resources. In this breakout session, participants heard from Koch AG, on how they leverage Amazon Monitron to implement predictive maintenance across their industrial machinery.

Chalk Talk session AIM240, “Bring the power of Generative AI to your employees with Amazon Q,” demonstrates how Amazon Q can provide secure, quick access to the power of generative AI for your employees. Amazon Q understands natural language, provides contextual answers using connected data sources, summarizes documents, generates content, and automates actions across enterprise applications and document repositories which can be especially useful to implement in worker guidance in shop-floor applications.

Supply Chain

AWS launched AWS Supply Chain service at re:Invent 2022. This service mitigates risks and lowers costs with an ML-powered supply chain application. Customers gain the end-to-end supply chain visibility necessary to track and trace the entire production process with unprecedented efficiency. The AUT207-INT Industrial Innovation session presented how cloud industrial companies across automotive, aerospace and consumer electronics leverage their data and cloud technologies to reinvent their business, optimize operations, accelerate time to market, and generate new revenue streams. Siemens presented how they use AWS to power their Xcelerator industrial software portfolio, new factory automation offerings, and how they built a virtual factory in the Industrial Metaverse. Honda talked about the collaborations with AWS in Japan and North America to accelerate innovation across product development, supply chain, and manufacturing.

Product Engineering

AWS empowers product developers and engineers to solve complex problems using High-Performance Computing (HPC) on AWS, model-based design, and large-scale parallel simulations. In the MFG 106 breakout session, speakers from Toyota Motor North America and Autodesk discussed how artificial intelligence can expedite computationally intensive simulations and modeling to accelerate product design and digital engineering, ultimately reducing time to market.

Sustainability and EV

In any electric vehicle (EV), the battery is the single biggest driver of sustainability and cost. The Chalk Talk session AUT 201, “Unlock the Power of Battery Digital Twins,” introduced a battery digital twin model—a virtual representation of a physical battery system. This session explored how Mahindra and Our Next Energy combine real-time vehicle battery data with machine learning algorithms, and data analytics to optimize battery performance, extend battery life, detect faults, and enhance safety.

Workshop IOT 305, “Detecting EV battery anomalies across a fleet using AWS IoT”, demonstrated a solution for early detection of electric vehicle (EV) battery anomalies, with participants gaining hands-on experience in managing, provisioning certification for a fleet of vehicles, working through vehicle modeling, campaign creation, and data ingestion and setting up a dashboard for insights.

In the demo area, AWS showed how our services are used by customers to help them build highly-scalable, low-latency OCPP EV charging CPO solutions on AWS. Another demo illustrated how AWS contributes to transparency and authenticity among stakeholders within a Battery Circular Economy. Additionally, AWS demonstrated how to optimize battery performance, extend battery life, and improve EV efficiency using Battery Digital Twins.

Conclusion

AWS presented innovation and customer success stories in eight strategic workload areas at re:Invent 2023. AWS has become the preferred cloud provider for Chinese OEMs expanding beyond the Chinese region, with announcements involving SAIC and BYD. AWS for Automotive teams continue to innovate with customers and on their behalf. Stay updated with other re:Invent 2023 announcements and explore AWS offerings on the AWS for Automotive page, or contact your AWS team today.

Related resources:
AWS automotive cloud solutions