Conference Program
Room A
Strategies, innovations and requirements for the safe deployment of ADAS and autonomous technologies
9am - 12.25pm
Moderator
9am
Deploying a safe and trustworthy AV in different markets
Now that we have removed the human from the vehicle has the AV become safer in traffic? No. We have removed the human and not the human element. But wait, is removing the human the solution? No. In fact the opposite is true, more the merrier.
In Vivetha's multi-pillared approach to continuously improve and deploy safe and trustworthy autonomous vehicles, 'inclusion is key'. Data, verification and personas need to be weighed equally. Humans in our different roles (as engineers, as drivers, as traffic inspectors etc.) are to be modelled using AI in order to understand and prove safety of AV.
What the audience will learn
- How to instill trust with end user to hail a Robotaxi
- How can safety of AV be measured using a multi-pillared approach?
- How sensible it is to share the AV Ecosystem?
9.25am
Collision avoidance bus system: radar and camera technology for safe transportation
Join us for an engaging talk on the revolutionary collision avoidance bus system developed by ZF, incorporating advanced radar and camera technology. This groundbreaking system aims to prevent collisions with cars, bikes, and pedestrians, ensuring enhanced safety on American roads. We will delve into the value propositions of this innovative driver-assist product and overall increased safety. Our discussion will also address the technological challenges associated with scaling up this autonomous product. Discover how this system is transforming the transportation and logistics industry, paving the way for a safer and more efficient future.
What the audience will learn
- Are you a bus manufacturer or MTA? Do you need a collision avoidance system?
- Are you a startup and want to collaborate?
- Do you need an intro into ZF?
- Do you need other hardware, software, or business help?
9.50am
Safely delivering autonomous trucking solutions
Torc is focused on helping the freight industry thrive by providing a safe, reliable, and cost-effective trucking solution. Delivering on safety requires going beyond just building a safe product. Collaboration with partners and stakeholders from across the ecosystem to inform the full lifecycle and customer journey is required. Attend our session to stay connected with how safety is shaping the future of freight.
What the audience will learn
- How Torc is approaching its’ safety management system to support a complex freight industry?
- What opportunities and challenges remain for the industry to deliver on safety?
- Why collaboration and sharing of safety information with partners and customers is paramount to achieving safety goals?
10.15am - 10.45am
Break
10.45am
Validator design and SoTIF processes to address the million miles problem
The “million miles problem” is not unique for AD design. It is an issue for ADAS too. Autonomous Emergency Steering and Autonomous Emergency Braking have the potential to cause harm and these faults can not be addressed by functional safety alone. Perception and decision algorithms which are insufficiently robust can lead to hazardous situations .
We will present validator concepts which can be used to design more robust ADAS as well as AD-systems. We will then also present how one can iterate the design to minimize the risk of functional insufficiencies and build an argument to motivate that the systems are safe.
What the audience will learn
- Why the million miles problem is relevant for ADAS
- The importance of tolerant time intervals in SoTIF
- Architectural design for handling the tolerant time intervals of functional insufficiencies
- How event scanners and tail distributions can be used in the verification and validation of ADAS.
11.10am
Cooperative congestion management
Traffic congestion is a major issue in many urban areas, resulting in increased travel times, air pollution, and economic costs. Congestion management is critical to improving traffic flow, but traditional approaches based on central control and infrastructure investments have limitations. Cooperative congestion management control is a promising new approach that optimizes traffic flow by coordinating driving speeds of connected and automated vehicles. Using a recurrent bottleneck on I-680 as a case study, this presentation will explore the key concepts, benefits, and challenges of the system.
What the audience will learn
- Overview of the congestion management
- Opportunities of connected and automated vehicles for solving congestion
- Concept of cooperative congestion management control
- Case studies of real world traffic
11.35am
Ensure Software Safety & Security With Continuous Testing
Agile methodologies foster a collaborative culture within organizations and have a proven track record of streamlining the product development process in the automotive industry. It places value on effective communication, integration, and cross-functional teamwork to enable continuous and rapid delivery of reliable products that are also safe, secure, and compliant to automotive industry standards like MISRA C 2023, ISO 26262, and ISO 21434.
What the audience will learn
- The high value of adopting a modern development CI/CD workflow
- Continuous testing: static analysis, unit testing, and code coverage
- TARA: Integrating security into your SDLC
12pm
Measuring surprising road user behavior
The key role of expectations and expectation violations (i.e., surprise) in the context of road traffic has long been acknowledged. Despite the important conceptual role surprise plays in traffic safety research, there is no precise quantitative definition or computational model of surprise in this domain. We demonstrate, for the first time, how computational models of surprise rooted in cognitive science and neuroscience combined with state-of-the-art machine learned generative models can be used to detect surprising human behavior in complex, dynamic environments like road traffic. In traffic safety, such models can support the identification of traffic conflicts, modeling of road user response time, and driving behavior evaluation for both human and autonomous drivers. We also present novel approaches to quantify surprise and use naturalistic driving scenarios to demonstrate a number of advantages over existing surprise measures from the literature.
What the audience will learn
- • We demonstrate, for the first time, how computational models of surprise rooted in cognitive science and neuroscience combined with state-of-the-art machine learned generative models can be used to detect surprising human behavior in complex, dynamic en
- • We present novel approaches to quantify surprise and use naturalistic driving scenarios to demonstrate a number of advantages over existing surprise measures from the literature.
- • We discuss the application of surprise in the identification of traffic conflicts, modeling of road user response time, and driving behavior evaluation for both human and autonomous drivers.
12.25pm - 1.35pm
Lunch
Room A
Software, AI, architecture and data management
1.35pm - 4.35pm
Moderator
1.35pm
Integrating general purpose and AD/ADAS-specific middleware
AUTOSAR Adaptive has a proven record as middleware for HPC nodes within next generation E/E architectures. AUTOSAR provides a range of services supporting development including diagnostics, communication, scheduling and security. However, as a general-purpose middleware it lacks features considered necessary for AD/ADAS development including reproducible behavior (for validation through simulation), a high-performance data transport layer (to support AD/ADAS bandwidth needs) and highest levels of safety (ASIL-D). In this presentation we briefly review the capabilities of AUTOSAR Adaptive and then consider how general-purpose and AD/ADAS-specific middleware can be integrated to combine the best of both when building an autonomous driving solution.
What the audience will learn
- The role and capabilities of AUTOSAR Adaptive as a general purpose Middleware component
- Requirements for AD/ADAS specific Middleware that are not covered by AUTOSAR Adaptive, including reproducibility, determinism and validation through simulation
- Integration points when combining Middleware components into a single platform
- What lessons have we learned from combining general purpose and AD/ADAS specific middlewares
2pm
Virtual ADAS/AV sensor validation in the cloud
The variety of ADAS/AV sensors that are available in the market today is big and growing. Automotive companies are not always aware of all existing sensor options, and even if they were, they would not be able to evaluate all the sensors in the real world because the process is long and expensive.
If at the end of the “real world sensor evaluation” process, the sensor performance is not satisfactory making it unsuitable for use in production, the automotive company needs to look for an alternative and start the evaluation process all over again resulting in delays and cost.
This session will focus on the benefits of “Pre-real-world sensor evaluation” using a cloud based platform for evaluating sensors virtually, comparing them and eventually narrowing down the number of sensors that will be taken for real world evaluation.
What the audience will learn
- AV/ADAS sensor types, and functionality
- The challenges of sensor selection for production programs
- Microsoft working with partners on ways to overcome these challenges by virtualization and cloud computing
2.25pm
Revolutionizing autonomous vehicles: harnessing the power of quantum computing for advanced automotive applications
Are you ready to experience the transformative potential of quantum computing in the autonomous vehicle industry? Join us as we delve into the groundbreaking applications of this emerging technology and its impact on the safety, reliability, and efficiency of self-driving cars. Discover how quantum-enhanced machine learning is revolutionizing object detection and data processing from advanced sensors like LiDAR. Learn how quantum computing's unparalleled processing speed can optimize route planning and reduce traffic congestion, enabling a safer and more efficient transportation ecosystem. This presentation will challenge your current understanding of the autonomous vehicle landscape and inspire you to envision a future where quantum computing drives innovation in the automotive sector. Don't miss this unique opportunity to explore the cutting edge of technology and its potential to reshape the world of autonomous vehicles.
What the audience will learn
- Understand how quantum computing enhances autonomous vehicle sensor data processing for faster and more accurate decision-making.
- Learn about real-world applications of quantum computing in improving route optimization and reducing traffic congestion.
- Discover the advancements in quantum machine learning for object detection and classification, contributing to safer autonomous driving.
- Gain insights into the challenges and limitations of integrating quantum computing technology into the automotive industry.
- Explore the future prospects of quantum computing in accelerating research and development in electric vehicles and novel automotive technologies.
2.50pm - 3.20pm
Break
3.20pm
Developing new ADAS and autonomous systems in the cloud
This presentation will look at some of the new development and deployment methodologies that can be used to develop new ADAS and AD functions. The SOAFEE initiative is taking cloud-native standards, tools and technologies and applying them to the development of software defined functions in the car. This allows for both early development in the cloud and then continuous integration/ continuous delivery (CI/CD) to the vehicle. We will cover some of the key technologies and methodologies in the SOAFEE architecture an show an example of how they can be applied to an autonomous software stack.
What the audience will learn
- How can cloud-native development methodologies can be used to develop and deploy ADAS and AD software
- What is the SOAFEE initiative and how to get involved
- How SOAFEE can be used to develop an autonomotous stack with a real-world example
- What are the key cloud-native technologies that can be used to develop vehicle software functions
3.45pm
Beyond scenario-based testing – validating the software-defined vehicle of tomorrow
Software, ADAS/AD and the user experience are the key differentiators for future vehicles. This presents new challenges to ensure the systems’ quality.
Developing advanced functions requires a complex, interdependent environment of different specialized tools and processes to satisfy the legal and technological requirements. Continuous integration and early detection of failures are critical for success. All system levels must be evaluated and validated. Scenario-based testing with variations only analyzes a system in an ideal state of operation.
This presentation will demonstrate how comprehensive test environments and state-of-the-art methodology achieve the highest level of software quality – continuously.
What the audience will learn
- - Challenges companies are facing to validate complex, software defined vehicles
- - How multi staging and DevOps helps to release high quality Software Features
- - What aspects beyond the method of scenario based testing have a critical impact on product success
- - Insights into the ideal test workflow from function developer to vehicle
4.10pm
Artificial Intelligence For AVs: A step-change in training data quality.
Artificial Intelligence (AI) is the only way a driverless vehicle can navigate dynamic public roads. But its effectiveness can only ever be as good as the training data it learns from. In this presentation, rFpro will provide insights into its new ray tracing rendering simulation technology, developed in partnership with Sony. It accurately replicates the way vehicle sensor systems perceive the world for the first time, significantly increasing the fidelity and quality of synthetic training data. Now available within High Performance Computing (HPC) solutions, it helps to remove one of the biggest challenges facing the industry; collecting high-quality, safety-critical training data for the development of AI.
What the audience will learn
- The benefits of using high-quality synthetic data to train artificial intelligence.
- The importance of physically modelling sensors, that allow for motion blur and rolling shutter effects to be accurately captured.
- How ray-traced synthetic data can overcomes the shortfalls of real-world only data collection.
- Benefits of decoupling simulation from real-time to overcome the trade-off between rendering quality and running speed. Available within HPC Solutions.
- How a technical partnership with Sony has created an End2End perception simulation pipeline for the training of artificial intelligence.
Room B
Standards, regulations and law, and their impact on engineers and technology
9am - 4.35pm
Moderator
9am
Regulations and standards for autonomous road vehicles worldwide: an overview
Surya KopparthiCertified Functional Safety Engineer & Autonomy Safety Professional
TÜV Süd America
USA
The worldwide legislative/regulatory and standards landscape for autonomous road vehicles is complex due to the following factors:
-Regional difference of legislation/regulations
-Multiple layers of legislation/regulations for a single administrative region
-Legislation/regulations leaving room for interpretation, including regarding applicability of industry standards
-Rapid changes in legislation/regulations
-Liability questions outside of legislation/regulations
This talk will give an overview of the achievements and issues in the field of legislation/regulations and standards for autonomous road vehicles.
What the audience will learn
- An overview of legislation/regulations for autonomous road vehicles worldwide
- An overview of industry standards for autonomous road vehicles
- An overview of liability in the context of autonomous road vehicles
9.25am
Test & deployment of autonomous vehicles in California – a regulatory update
California is the birthplace and epicenter of autonomous vehicles. And a regulatory structure governing the technology has been in place for almost 10 years. Automotive and technology companies are testing their autonomous vehicles on the public streets in California. Some companies have deployed their vehicles in a variety of business models. Learn about the latest developments, who is testing where, and what companies have deployed their vehicles, plus lessons learned in the last 10 years. Then discover what engineers need to consider when testing & deploying the next generation of autonomous vehicle technology in California’s evolving regulatory landscape.
9.50am
The evolving regulatory environment for Advanced Driver Assistance Systems (ADAS)
It has been two years since the Bipartisan Infrastructure Law (BIL) was enacted and the National Highway Traffic Safety Administration (NHTSA) has issued Notices of Proposed Rulemaking (NPRM), initiated research in support of future regulatory actions, and moved forward with required updates to the New Car Assessment Program (NCAP). This presentation will address the ADAS technologies the BIL requires NHTSA to address, and what the latest NHTSA actions are in the regulatory and research stages in support of these required actions. Insights to NTHSA's future actions will also be included.
What the audience will learn
- The ADAS technologies affected by the Bipartisan Infrastructure Law (BIL)
- NHTSA's latest regulatory and research activities to address the BIL requirments
- Insights into expected future actions by NHTSA related to ADAS and ADS
10.15am - 10.45am
Break
10.45am
AV privacy and cybersecurity: legal and regulatory framework and considerations
This presentation will explore the current legal and regulatory landscape surrounding autonomous vehicle privacy and cybersecurity in the US and beyond. Attendees will learn about the various privacy and cybersecurity risks as well as strategies to manage those risks.
What the audience will learn
- US Federal laws and regulations and recent/upcoming developments
- US State laws regulations and recent/upcoming developments
- Laws Regulations and recent/upcoming developments outside the US
- Privacy and Cybersecurity Risks in Autonomous Vehicles
- Strategies to Manage Privacy and Cybersecurity Risks in Autonomous Vehicles
11.10am
Product Calculus Series - Wireless communications and ADS Safety: The state of V2X regulatory developments.
Stakeholders in the mobility and transportation industry have long envisioned vehicle safety communications enabled by direct communication technology (V2X). Yet this vision has enjoyed limited deployment due to many challenges in getting V2X technologies to market. This presentation tackles one pervasive but necessary challenge: regulation. Specifically, what is the state of V2X regulation in the US, and how does this evolving framework factor into your product calculus for ADS technology?
What the audience will learn
- • State of V2X regulatory development in the United States
- • How US federal agencies are collaborating to regulate V2X technologies
- • Role of the public and private sectors in getting V2X technologies to market
- • Complexities presented by current regulatory uncertainty
11.35am
AAMVA Guidelines for Testing and Deployment of Automated Vehicles
Paul SteierDirector, Vehicle Programs
American Association of Motor Vehicle Administrators (AAMVA)
USA
This session will provide a detailed overview of the AAMVA publication titled Safe Testing and Deployment of Vehicles Equipped with Automated Driving Systems Edition 3. This guidance document that will be reviewed provides assistance and recommendations for the regulation of testing, piloting, and deployment of automated vehicles. As there are no national standards for regulating automated vehicle operations, this guidance serves as a benchmark for oversight of safe testing and deployment of these vehicles. This overview will provide an in-depth discussion on guidance given for the main areas of consideration consisting of the vehicle, operator, and law enforcement.
What the audience will learn
- Undertand the role and responsibilities states have taken on with testing and piloting of AV
- Gain an in-depth understanding of the AAMVA Guidance document
- Provide input for future AV testing and piloting for state AV regulatory guidance development
- Learn about law enforcement involvement and considerations with AV use
- Obtain contacts to collaborate with state regulators
12pm
A vision for promoting AV rules & public confidence through clarity in safety engineering standards
AVs are caught in a chicken-and-egg problem: Regulators want rules to wait for increased public trust, while improved trust depends on compliance with regulatory standards.
This presentation offers a solution to this logjam: First, Different organs of government can take steps to apply existing motor vehicle approaches to AVs. Second, third-parties can encourage developers to compile a clear, unitary description of their safety approach. By providing a public accounting for safety, this approach can start providing regulators with a firmer foundation and richer, more technical, more developed content that can serve as a foundation for new AV-specific rules.
What the audience will learn
- Safety standards are not moving forward with any speed. This dynamic results, at least in part, from a chicken-and-egg problem: Regulators want rules to wait for increased public trust, while improved trust depends on compliance with regulatory standards
- The absence of safety standards directly affects further development: Industry is left with little certainty if regulators could, at any moment, dramatically change design expectations.
- Even in the absence of total clarity on ultimate safety standards for AVs, government regulators in different organs of government can take steps to apply existing motor vehicle approaches to AVs, and thereby steadily fill in our collective understanding
- Outside government, third-parties can help encourage movement toward greater regulatory certainty. Including by encouraging developers to compile a clear, unitary description of their safety approach, and the development of industry standards
12.25pm - 1.35pm
Lunch
Moderator
1.35pm
AI components: technology, validation, liability and solutions for ADAS/AI
AI components are a powerful new capability which is the center of ADAS and AV solutions. However, AI components present significant challenges in the domain of validation and verification. Without a clear methodology to handle AI components, ADAS/AV solutions introduce significant legal liability to product developers. This presentation will present the history of AI components, the V&V challenges, the impact on ADAS/AV legal liability, and structural pathways towards successful integration. Without a comprehensive approach to AI components, the ultimate value for functionality such as ADAS is limited, but with reasonable solutions, the value perceived by customers rises rapidly.
What the audience will learn
- History and Role of AI components
- Validation Challanges for AI components
- Validation Challanges for ADAS systems with AI components
- Legal Perils with current V&V OEM Methodology
- Pathways to successfully navigate ADAS validation and manage legal liability
2pm
Open Standards: Making sure they actually work
Open standards make it easy to exchange data between tools and stakeholders. At least that is the goal, but the reality is often different. As adoption of ASAM OpenX standards in the industry grows, we observe that there are often differences in interpretation or implementation of the standards. This often means that it is difficult to determine whether a specific tool or file actually conforms to a standard. Standards need to be supported by test suites, checker tooling and more to really enable consistent understanding of the standards. This talk will give some insight into the latest developments at ASAM on the journey to deliver more than 'just' standards.
2.25pm
The future of global AI governance
Peter StockburgerSan Diego Managing Partner and Co-lead Global Autonomous Vehicles Practice.
Dentons
USA
Governing autonomous systems presents a significant challenge when taking into account the needs of all stakeholders, including engineers. While the demands for governance are clear, the path forward remains uncertain. Peter Stockburger from Dentons, and Dan Richardson from the Spatial Web Foundation, will be presenting next generation socio-technical standards and related policy proposals that will help drive global discourse on the governance of Autonomous Systems. These standards have the potential to address what current regulatory and governance efforts cannot: the critical issues of interoperability, explainability, and AI’s exponential advance toward greater intelligence and autonomy.
What the audience will learn
- An overview of the socio-technical standards being developed by the Spatial Web Foundation in partnership with the IEEE
- How Socio-technical standards ensure interoperability of data, models and systems.
- How Socio-technical standards make it possible for Autonomous Intelligent Systems to be compliant with diverse local, regional, national, and international regulatory demands, cultural norms, and ethics.
- About the
2.50pm - 3.20pm
Break
3.20pm
Exploring the latest updates of MISRAC:2023
Learn about the latest updates in MISRAC:2023, the widely used coding standard for embedded software development, and how these updates specifically apply to the automotive industry. Through practical examples and case studies, we will discuss key changes in MISRA C: 2023, highlight their implications for automotive software development and share best practices for implementation. Stay ahead of the curve and ensure compliance with industry best practices for developing safe and reliable embedded software in the automotive domain.
What the audience will learn
- What's new in MISRA C: 2023
- Example cases on how the new rules work
- How MISRA C: 2023 affects the automation industry.
3.45pm
Legal and consumer requirements for ADAS in US and Europe
The deployment of ADAS in the American and European market is guided by legal and consumer requirements defined by local stakeholders. Currently, US defines consumer requirements under NHTSA US NCAP programme and a new set of protocols is available. In Europe, the General Safety Regulation defines mandatory requirements that urge for standard fitment of multiple ADAS in all new vehicles. And beyond the legal framework, Euro NCAP foresees complex safety and assisted driving systems, with a comprehensive roadmap until 2030.
The presentation will provide review current and future requirements for ADAS applicable to the American and the European markets.
What the audience will learn
- Current and future NHTSA US NCAP protocols for ADAS
- Legal framework in EU defined by the General Safety Regulation with focus on ADAS
- Overview of the Euro NCAP Vision 2030 roadmap for safety and assisted driving systems
- Overview of legal and consumer requirements applicable to the American and European markets
4.10pm
Complying with UNECE R155 and R156
The automotive industry continues to evolve towards a software-centered and -defined vehicle. To ensure automotive safety in this new world, legislation is coming into force through the work of the UNECE Working Party / WP 29.
While these new requirements can seem daunting at first glance, they are a formalization of best practices for over-the-air (OTA) software updates and cybersecurity controls. Still, missing key requirements or uncertainties around implementation can lead to unnecessary compliance failures, delays, and wasted time and money.
We will go over the key requirements, how they relate and a practical implementation of compliance.
What the audience will learn
- An overview of ISO/SAE 21434, UNECE R155, R156 and how they relate
- Key requirements dictated by the standard and regulations
- Outline of a general-purpose compliant OTA update architecture
- An example of a concrete implementation using Mender
Room C
Advanced simulation and scenario-based testing
9am - 4.35pm
Moderator
9am
Smart & efficient methods for testing AVs
Autonomous vehicles are a key enabler for increasing safety and access to mobility for current society, but they can also strongly contribute to the CO2 footprint of current vehicles, by an efficient driving behavior. To enable AVs on mass scale, it is crucial to be able to have smart and efficient testing techniques that assure proper operation of the system in all environmental conditions. This presentation will give an insight on such approaches that make use of ODD information, AI- and combinatorial based testing for enabling a comprehensive test program with highest coverage of the parameter.
What the audience will learn
- - How to address the issue of coverage for testing AV systems
- - The multi pillar approach used for creating the safety argumentation with respect to verification and validation process.
- - Ontology-based test generation method, which makes use of the ODD and combinatorial testing to automatically generate efficient test programs
- - Game-based method for creating critical test scenarios by making use of gaming industry concepts for generating edge and corner cases.
9.25am
A quality of exposure metric for SOTIF validation
The question of how many miles is enough to determine readiness for driverless deployment has been a popular topic in the automated vehicle industry for some time. This presentation aims to explore the various aspects that should be considered in the decision-making process to establish driverless readiness. While a specific mileage accumulation target is perhaps the easiest to establish as a concrete goal, the oversimplification of this approach and associated fallbacks are laid out here. As a result, a comprehensive approach is required to establish such a target. One new concept Motional is using is the Quality of Exposure metric.
What the audience will learn
- There are many issues with using a pure quantity of miles to validate AV performance
- True AV performance depends on many factors including capabilities or features exercised
- Highly accelerated life testing can be used to argue for lower validation miles by
- Statistical methods can be used to analyze the number of miles to build a certain confidence
- Unique metrics can be applied to validate AV performance against human benchmarks for certain behaviors
9.50am
Leveraging ODD at VW Commercial Vehicles: ODD specification, validation and coverage
Dr Schwalb will provide an overview of the components of the ODD specification. The presentation will demonstrate how a detailed ODD can be developed in a modular distributed fashion whereby different aspects (e.g. highway pilot vs parking assistant) can be integrated into a single coherent interpretable specification. We will review how the Operational Domain (OD) is leveraged, how coverage is determined and clarify the relationship to the scenarios used for testing. For each use case, the applicable OpenODD aspect will be highlighted and references will be provided to more detailed method documentation available in the form of IEEE publications from the group.
10.15am - 10.45am
Break
10.45am
Testing ADAS and AVs in a deterministic simulation environment
Its essential that we can test our ADAS and AV systems on a wide range of scenarios covering nominal, critical and edge cases. This must be done in a deterministic simulation environment so that the same test can be run in repeatable manner with the only variable being the ADAS/AV controller behaviour. AVSandbox provides a deterministic and sensor realistic simulation environment that allows the ADAS/AV controller to be immersed into the virtual world and tested on a diverse range of scenarios to identify failure modes and focus the development effort to improve safety.
What the audience will learn
- Insights into the challenges of virtual testing and how to effectively harness this to drive development
- Why deterministic simulation is important for ADAS and AV system development and testing
- How a sensor realistic simulation environment enables the ADAS/AV controller to be immersed into the virtual world
11.10am
Safety-driven validation – enable large-scale ADS/ADAS safety validation using ASAM OpenSCENARIO 2.0.0
The presentation describes a new approach to large scale safety verification and validation of ADS/ADAS. The approach relies utilizes ASAM OpenSCENARIO 2.0.0, coupled with scenario-based coverage-driven validation, producing safety metrics. This approach changes the way the automotive industry is conducting verification and validation of automated driving systems. It also demonstrates an advanced method for safety verification that allows tackling the infinite space of scenarios. Several use cases will be presented, with an emphasis on how the new approach enables a step function in ensuring the correctness and safety of automated driving systems.
What the audience will learn
- A full flow/solution for Safety driven validation – tools and methodology and their usage.
- Usage of ASAM OpenSCENARIO® 2.0.0 advanced features for safety scenarios.
- Producing metrics, data and evidence for safety argumentation.
11.35am
Accelerating autonomous driving development through simulation-based validation and verification
The presentation will focus on simulation-based validation for autonomous driving, which has emerged as an effective tool to complement physical testing and accelerate the development process. We will discuss the benefits and challenges of simulation-based validation for autonomous driving, including the use of advanced sensor models, scenario generation tools, and real-time simulation platforms. Additionally, we will present the latest trends and innovations in the autonomous driving simulation and validation field, along with case studies and examples of successful simulation-based validation projects. Attendees will learn how to leverage simulation-based validation to ensure the safety and reliability of autonomous driving systems.
What the audience will learn
- Benefits of simulation-based validation in addressing the challenges of autonomous driving.
- The techniques and tools available for simulation-based validation, such as advanced sensor models, scenario generation tools, and real-time simulation.
- The latest trends and innovations in the autonomous driving simulation and validation field.
- Real-world examples of successful simulation-based validation projects
- Best practices for leveraging simulation-based validation to ensure the safety and reliability of autonomous driving systems.
12pm
Critical scenario creation methodology for safety assessment
The design of infrastructure can have an impact on overall safety and in most accident cases, is also attributed to the driver. The introduction of the SOTIF standard has brought along a fundamental shift in the safety case for autonomous vehicles resulting in the vehicle manufacturer carrying an increased burden of responsibility and liability.
Critical Scenario Creation (CSC) is a proprietary methodology to systematically and automatically generate unsafe-unknown scenarios. By adopting the CSC process, cities and traffic planners can identify problem areas and evaluate infrastructure changes that support safer operations, thus increasing safety in the urban environment.
What the audience will learn
- How to automatically generate Critical Scenarios and asses safety in context of ADAS and AV functionality?
- Generation of scenarios to evaluate safety per SOTIF standard?
- How to evaluate the safety impact of Infrastructure changes?
12.25pm - 1.35pm
Lunch
Moderator
1.35pm
Digital Twin Concepts for Autonomous Vehicle Testing
The pace of innovation and digital transformation in the automotive industry is growing exponentially. The push toward software-defined vehicles and digital twins enables engineers to accelerate the definition, design, and production of cars, modules, and critical sensors for the automotive industry.
The environment remains a variable when validating the functionality of a new vehicle in the real world. This context is beyond the control of automakers, creating a challenge when testing a car’s functionality with confidence. However, automotive manufacturers do have control over testing in the lab using digital twins. In fact, when testing, you can analyze, record, and assess
What the audience will learn
- Facilitate the path to autonomy with in-lab testing.
- How to validate radar-based Advanced Driver Assistance Systems (ADAS)/ autonomous driving (AD) with 512-pixel resolution.
- Synchronize with and prove V2X functionality.
- Integrate this testing with any hardware-in-the-loop system, 3D modeler, or V2X stack software.
2pm
Measuring digital impact on the zero-crash paradigm
This presentation centers around the integration of the virtual and simulation components within the AV and ADAS domains, aiming to achieve a zero-crash paradigm. Using virtual technologies with a blend of simulated and real-world situations, vehicles can undergo comprehensive testing that enhances their ability to respond and intervene effectively in critical contexts. Through simulation and validation of a wide array of driving scenarios, preemptive crash avoidance strategies can be swiftly deployed. This approach permits exhaustive scenario validation in the virtual realm, reserving physical testing for unique edge cases, thus optimizing resource allocation, value, and time-to-market. The ultimate outcome manifests as more resilient AI, AV, and ADAS systems, yielding a substantial reduction in accidents, injuries, and fatalities.
2.25pm
Accelerating the data loop
The physical world is complex and highly dynamic, with sudden changes and unforeseen anomalies. Therefore, it is essential to establish a systematic way to prepare autonomous systems for safe operation in the physical world.
In this presentation, we give insights into our novel AI Scenario Engine that helps customers deploy autonomous systems significantly faster and with dramatically less cost. Core of this platform is a world-class computer vision software that extracts traffic scenarios from monocular cameras in a highly accurate and fully automated way.
DeepScenario's AI Scenario Engine has been used by the world’s most recognizable companies, like BMW, Bosch, or Torc Robotics, to solve the challenges of autonomous system deployment.
What the audience will learn
- Conduct large-scale traffic observations significantly faster and at dramatically less cost with stationary cameras.
- Gain knowledge from motion data by parameterizing the real-world measurements.
- Convert critical scenarios into industry standards to enable continuous training and testing in simulation.
2.50pm - 3.20pm
Break
3.20pm
Workflows for generating virtual scenarios from recorded vehicle data
This presentation introduces workflows for generating virtual scenarios from vehicle logs utilizing Automated Driving Toolbox™ and RoadRunner Scenario.
The workflows consist of multiple data processing tasks, including:
• GPS and IMU sensor fusion to generate ego trajectory
• Ego localization using lane detections and a high-definition map
• Target vehicle trajectory creation from recorded lidar, radar, and vision sensors data
RoadRunner Scenario exports the scenarios created to OpenSCENARIO, which is used for regression testing of the advanced driver-assistance systems (ADAS) and autonomous driving (AD) algorithms.
What the audience will learn
- Create virtual scenarios from recorded vehicle data
- GPS and IMU sensor fusion to generate ego trajectory
- Ego localization using lane detections and a high-definition map
- Target vehicle trajectory creation from recorded lidar, radar, and vision sensors data
- Export the scenarios to OpenSCENARIO
3.45pm - 4.35pm
Panel Discussion - Safety validation for Highly Automated Driving
Highly Automated Driving offers many potential benefits. But despite technological progress, timelines for large-scale commercial deployments remain uncertain, the main barrier being safety. In this panel, experts from the industry will discuss and provide insights regarding the challenges of safety verification and validation and the innovations needed for assuring safety and building a valid Safety Case.
