Synopsis: For automated driving systems (ADS) to be successfully adopted, enhancements to the road infrastructure in the form of better lane markings and signage have limited impact. Being upgraded with advanced sensing technologies, however, the infrastructure is capable of playing a much more significant role, making the transportation intelligent by actively assisting the ADS. By becoming smart, infrastructure will provide all road users with accurate situational awareness, making sure everyone is on the same page, speaking the same language. Not only will such digital infrastructure provide the information needed for safer operation of ADS, it will also improve their effectiveness.
co-founder and COO
Synopsis: Using high-resolution radar for autonomous driving poses new challenges and brings new opportunities in the mapping of the environment, including the tracking of other cars and pedestrians, and localizing the self-driving car within its environment. Oz will discuss the role of advanced algorithms with next-generation radar, which senses the road with both an ultra-high resolution and a wide field of view. He will focus on how post-processing and SLAM algorithms can help resolve ambiguities and achieve low false alarm rates.
director, automotive solutions and platforms
Synopsis: As companies strive to build Level 3+ autonomous systems, they are realizing that the vast amount of compute power required to perform the autonomous functions is not really practical for mass production due to the size, power consumption and thermal properties of today’s computing platforms. This talk will investigate options to build more efficient autonomous vehicle compute platforms to conform to the strict power requirements required by EV and hybrid vehicles, while still delivering the necessary performance to manage the autonomous decision and action process.
Synopsis: The presentation will discuss: lightweight HD mapping with cameras, human-guided AI for map making, 3D maps enabling full autonomy, and use of 3D maps as content in ADAS simulations.
co-founder and VP of engineering
Cepton Technologies Inc
Synopsis: Autonomous vehicles (AVs) will revolutionize transportation around the world. Attendees will learn why high-resolution, low-cost lidar is key to mass deployment of AVs around the world in a safe and cost-effective manner. They will also learn about new advances in 3D lidar sensing technology and its application in AVs. This talk will focus on the new development in Cepton’s lidar technology and the future technology roadmap. And how Cepton’s high-performance Vista lidar’s long range and resolution help with object detection and perception in challenging environments, enabled by the computing power of GPUs.
director business development
Synopsis: The presentation will discuss the security of digital smart grid control of connected autonomous vehicles; identification of challenges in disruptive technologies and their effect on operations and communications of ITS control systems; safety and reliability of embedded software in autonomous vehicle control and sensor systems.
Synopsis: In this session VSI and Dataspeed will detail the steps in building an L2+ automated vehicle. Through its own advanced build, VSI will discuss system performance requirements as well as the necessary middleware to support L2+ automation including sensor fusion, precision localization, over-the-air updates and driver monitoring. Dataspeed will discuss the technicalities of by-wire control systems, which are the foundation of most self-driving cars. Delegates will come away with a deeper understanding of how a self-driving car is built, the challenges and the gaps.
senior applications engineer
Synopsis: Different virtual validation use cases require different sensor models. Accuracy, simulation performance and usability play major roles. This presentation will give an overview of the different dSPACE solutions for sensor simulations. In this presentation you will learn: how to speed up simulations using ASM (automotive simulation models) ground truth sensor simulations; how to generate raw data streams for environment sensor simulations such as camera, radar and lidar sensors using GPU-based sensor models in MotionDesk; new solutions for HIL testing of highly automotive driving features involving radar, camera, etc.
manager, ADAS and cybersecurity
Synopsis: Autonomous driving, ADAS and connected vehicle applications have flooded the vehicle with features, exponentially increasing the attack vectors in the vehicle and the need to test them. Autonomous enablers such as cameras and connectivity enablers such as wi-fi or Bluetooth can be tested at a component level with dedicated security bench setups. The challenge is to test for security at a system level as opposed to only component-level security testing. A simplified approach includes identifying vulnerabilities, reducing risk by resolving threats from the identified vulnerabilities, and validating the existence or non-existence of known and unknown vulnerabilities within the target system.
Ford Motor Company
Synopsis: The automotive industry is heading toward the path of autonomy with the development of autonomous vehicles. As in-vehicle testing for autonomous vehicles will be considered expensive, time-consuming and unsafe due to the number of scenarios and driven kilometers required for validation, a simulation platform that can provide a controlled and consistent testing environment is required for rapid prototyping and testing of the autonomous vehicle. This paper focuses on a powertrain and chassis hardware-in-the-loop (HIL) simulation of the autonomous vehicle platform and the correlation of the performance of the corresponding subsystems with those of the actual autonomous vehicle.
founder and CTO
Synopsis: The presentation discusses and analyzes the current status of AV verification, using recent examples. It analyzes challenges to eventual deployment, noting that we can expect many fatal AV accidents. It then suggests that a comprehensive, transparent verification system could help solve this inevitable tension. Finally, it describes principles of verification using a scenario-based, coverage-driven methodology.
Synopsis: Level 2+ autonomous vehicles require a fusion of perception sensors (lidar, camera, radar) and absolute position sensors (GNSS, map, IMU) for safe operation. This presentation will explain how GM performs QA activities and pre-production validation of GNSS and map accuracy.
director of product management, telematics
Harman, a Samsung company
Synopsis: By 2025, the number of connected things is expected to grow to 1 trillion. It is this deluge of new devices that will demand a paradigm shift in the network’s capacity to handle the devices and the data that will be generated by them. The technical requirements that necessitate a true generational shift from 4G to 5G are sub-1ms latency and downlink speed greater than 1Gbps. Smart, connected and autonomous driving will require a robust and omnipresent wireless network that has extensive coverage, high data transfer speeds, ultra-low latency and ultra-high reliability – qualities that can only be found in 5G.
senior principal engineer and the chief systems architect for automated driving solutions
Synopsis: There is little argument that machines will be better drivers than humans. Yet there is very real risk that self-driving vehicles will never realize their life-saving potential if we can’t agree on standards for safety. We will explain how RSS provides specific and measurable parameters for the human concepts of responsibility and caution, and defines a 'safe state', where the autonomous vehicle cannot cause an accident, no matter what action is taken by other vehicles. We will also talk about how the industry can collaborate to help put these types of safety standards in place.
International Transportation Innovation Center
Synopsis: The world of transportation is rapidly changing through digitization, leading to a portfolio of smart and sustainable mobility services from which the user can choose on demand and – due to full automation – without the need to be able to drive a vehicle. How those multi-modal mobility services interact with each other and with the supporting infrastructure has to be validated in cyberphysical testbeds in smart city environments to determine the operational safety and security risk level. This presentation will discuss design criteria for closed and open testbeds that could support a certification process for automated driving.
Isuzu Technical Center of America Inc
Synopsis: Global deployment of autonomous capability for commercial vehicles is a big challenge. In order to improve the robustness of autonomous approach under different traffic scenarios, environments, road conditions and driver behavior, a combined approach of physical testing, simulation, HIL and on-road testing has been established for sensors (camera, radar, etc.) and algorithm verification. Virtual testing is employed to reduce dangerous moving-target physical tests; a conversion method has been developed to utilize databases from different resources; machine learning is used to identify the worst-case scenario in the combined database.
Kanagawa University, Japan
Synopsis: We have tested Comma.ai’s autonomous driving system during a round trip from Tokyo to Osaka, which are the two biggest cities in Japan. We drove successfully for nine hours on the outward journey, but experienced two system failures during the return journey. This was caused by transforming analog car data to digital data through a USB interface. Except for this point, the present system is reliable on driving one lane of Japanese highway. We will publish multi-camera recoding data from this monumental 18 hours of driving. A remaining problem is to build a system that can manage branching and merging of lanes.
director, global business development, ADAS
Synopsis: As automotive sensors increase in design complexity to deliver multiple capabilities, their test routines have evolved to complex test protocols in order to verify functional performance of the sensors. Radar, lidar and camera sensor technologies form a quintessential part of an autonomous vehicle mapping its environment before deciding on required actions. The combined use of scenario-based testing and measurements can effectively verify functional performance of the sensors in their application use case environment or conditions. In this session, we will share details of how the ADAS iiT sensor fusion HIL test method can address this evolving need.
strategic innovation director
Synopsis: What are the challenges and some solutions for user (citizens and cities) adoption of autonomous public transportation? What are some key topics about safety and how to address them? Based on experimentations in Europe with government and counties, such as SCOOP, OPERA projects and autonomous shuttle in Nantes city, we will review these challenges and some tested solutions. Topics such cybersecurity, the role of infrastructure for safety, proactive information from sensor through V2X, and best practices for adoption by citizens and cities will be addressed, supported by real-life examples. The purpose is to enrich rules and regulations with facts.
senior electrical engineer
Synopsis: Cooperative automation has the ability to improve highway capacity and traffic flow. This has been demonstrated in simulation for a managed lane scenario with access limitations (i.e. lanes open to CAVs and connected vehicles only), as might be expected for early deployment on managed lanes. These applications were then tested on a real managed lane to confirm those benefits. The cooperation between multiple stakeholders with different operational objectives made this testing successful. This will be a review of the inception, planning, execution and special considerations of testing on this public facility and potential early deployment area.
chief software engineer
Synopsis: The Federal Highway Administration (FHWA) conducts research in connected, automated vehicle operations. CARMA is a reusable and extensible vehicle control middleware that provides an API enabling researchers to easily test various vehicle guidance algorithms, and supporting vehicle-to-vehicle and vehicle-to-infrastructure interactions. CARMA facilitates cooperative automation. These behaviors involve negotiation between vehicles to safely achieve workable solutions for spatial conflicts. FHWA has defined a new set of dedicated short-range communications mobility message types, extending SAE J2735 to support negotiation between vehicles about future intentions. Our goal is to see these new message types included in a future version of the J2735 standard.
Messring Systembau MSG GmbH
Synopsis: Pedestrians and cyclists account for a significant proportion of road deaths worldwide. Current ADAS test systems are tackling this challenge, but are limited in their design to linear or two-dimensional motion. With this setup, particularly during acceleration processes, an unrealistic motion is generated. The concept of hanging dummies from above creates new possibilities for more life-like dummy trajectories using six degrees of freedom. The system sets new standards in precision and repeatability through the ability to reproduce real-life human motion sequences and imitate them realistically – for example, based on data from a motion capture system.
head of business development
Synopsis: Traditional automotive testing tools are not up to the task of, or able to scale to, the level needed to satisfy safety requirements for autonomous vehicles (AVs). Simulation that encapsulates mature, agile software engineering approaches, especially those involving continuous test and integration, provide a proven way forward. These approaches provide endless miles of virtual testing needed for validation in a single cycle, outpacing physical testing by an order of magnitude. Attendees will learn about current safety regulations for AVs, general AV testing best practices/methodologies and novel AV simulation approaches that have the potential to satisfy auto makers, policymakers and consumers alike.
Synopsis: Safety is not an aftermarket feature. We present quantitative methods for baking in safety, and achieving reliable integrated drivers using unreliable AI components. We explain how dense safety rewards can be leveraged by reinforcement learning methods to achieve safe competent driving. Finally, we discuss an approach for continuous integration capable of covering hundreds of millions of miles per hour.
principal R&D engineer
MTS Systems Corporation
Synopsis: Conducting AV/ADAS testing on public roadways is highly problematic in terms of safety and repeatability. Simulation presents a safe and repeatable alternative, yet falls short in capturing the subtleties of real driving scenarios. To overcome these challenges, MTS is combining virtual and physical elements – models, physical components, sensors and humans – into hybrid simulation environments, bringing new levels of safety, repeatability, fidelity and efficiency to vehicle development. This presentation explores the application of hybrid simulation for AV/ADAS test and development, covering conceptual models, the blending of virtual and physical, repurposing existing machines and models, and product validation and verification.
business development manager - Americas Automotive
Synopsis: Autonomous vehicles require complex controls that must be proved to be safe and reliable. From ADAS interfaces to ride-sharing applications, performance requirements are ever increasing. A test workflow to verify designs must be quick to implement, operationally effective, and comprehensive in quality and performance. A hardware-in-the-loop test workflow provides the efficiency of simulation with the completeness of performance testing in a practical test budget and schedule. This presentation provides an overview of the test needs and demonstrates a HIL verification workflow to satisfy them. The audience will learn a simple approach to test autonomous vehicle controls.
senior director of market development, autonomous vehicles
Real-Time Innovations (RTI)
Synopsis: Autonomous vehicles are the most disruptive change in automotive history. Car companies are scrambling to keep up with the investment and technology requirements, but the future is still unclear. How will this technology be deployed? When? By whom? There is no common accepted standard, no dominant platform and many unanswered technical questions around safety and security. Without solving these challenges, we cannot have a robust industry ecosystem. Who will save us from this confusion and indecision? Enter the DDS Standard. The industry needs a common framework that can meet the performance, safety and security requirements necessary for Level 4.
Synopsis: Training and testing artificial intelligence algorithms (deep learning neural networks) supplemented by synthetic simulated sensor data improves performance and adds to the testing approaches. The sensor models used include camera, radar and V2X, with appropriate segmentation. These models can be used to produce ROC (receiver operating characteristic) curves and other measures of detection and estimation system performance. Examples of the process are offered in this presentation.
president and CEO
Sirab Technologies Inc
Synopsis: Although many autonomous vehicle approaches are being pursued globally, 'safety assurance' with highly reliable guidance systems is still a key challenge. The Sirab team, with core expertise in building safety- and security-critical systems for aviation, is trying to solve this challenge with our unique patented approach that can support in the deployment of a platooned convoy of commercial vehicles at high speed to improve safety, capacity and operational efficiency of road transportation systems. Our core technology consists of a highly reliable guidance system using radar and a modular architecture for lane keeping and platooning, enabling solutions with safety assurance.
senior business development engineer
Spirent Communications PLC
Synopsis: Time sensitive networking protocols and applications are the enabler for using real-time functionalities in different industries based on Ethernet/IP networks. For the automotive world, TSN will help to implement driver assistance and autonomous functions into next-generation vehicles. This presentation illustrates use cases of time-sensitive networking in vehicles; it includes ways to validate network components like ECUs and end devices such as cameras and sensors as well as software applications using this technology. Delegates will hear how autonomous vehicles act like data centers and learn how to avoid safety issues based on malfunctions or cybersecurity attacks.
executive director, Automotive Grade Linux
The Linux Foundation
Synopsis: The race to roll out new technology features and autonomous vehicles continues to heat up. In order to compete at the speed of a tech company, many auto makers have shifted from traditional development processes to agile, rapid development through open-source software. Dan will provide an overview of AGL, key milestones and the project roadmap. He will also discuss AGL's vision for functional safety as well as for an open-source platform for autonomous driving that will help accelerate the development of self-driving technology while creating a sustainable ecosystem that can maintain it as it evolves over time.
Frans de Rooij
director business development
Synopsis: Vehicle sensors, such as cameras and radar, have powered the introduction of advanced driver assistance systems. They need to be combined with a high-definition map (HD map) to make higher levels of driving automation safe and comfortable. We will discuss how the data layers from TomTom’s HD map are correlated with vehicle sensor data to enable accurate localization, environment perception and path planning. We will also show how the sensor-derived observations ('Roadagrams') contribute to keeping the HD map up to date, and how the updated map is streamed to the vehicle.
Synopsis: Although there is little to no doubt that full automation will ultimately be a reality, we must ensure that its deployment occurs in a safe and responsible way. From a regulator’s perspective, safety should always remain the paramount priority. As ADAS are leading the way and becoming the norm, we should also ensure that marketing does not mislead consumers and that the limitations of various systems be communicated clearly. The results of an in-depth analysis of key ADAS features from 30 commercial systems will be presented, and the main safety benefits and concerns associated with partial automation will be discussed.
U.S.DOT Federal Highway Administration
Synopsis: CARMA monitors vehicle speeds to facilitate cooperative tactical maneuvers among all cooperative automated driving systems (CADS). CARMA is open source and easily shared and integrated. The Federal Highway Administration (FHWA) developed the innovative platform to encourage collaboration with the goal of improving transportation efficiency and safety through early adoption. CARMA will be released on GitHub for testing. The unique platform supports vehicle automation capabilities across multiple vehicles and vehicles types with security features. Beyond reducing traffic congestion and improving transportation safety, CARMA decreases the risk for application research and development (R&D), expands on existing automation capabilities and reduces R&D time.
senior product manager
Synopsis: In this talk we will discuss approaches to software performance measurement using scenario-based simulation. These approaches are intended to shed light on an industry-wide problem: the lack of a gold-standard 'driving test' for self-driving software. The presentation will cover: road-based performance metrics, scenario framework, the role of simulation in self-driving software, and a proposal for scenario-based metrics in simulation. Leveraging simulation as a validation method and aligning it with published scenario testing frameworks will accelerate the evaluation of mass-market self-driving vehicles.
Synopsis: The standard of quality in industry-accepted benchmark data falls short of the quality required for Level 4 and 5 autonomous driving. In this session we highlight research findings and best practices to improve annotation for training and validation data at scale. Mobility companies can now differentiate based on the quality of their training and validation data.
advanced development lead
Synopsis: Autonomous driving Levels 3-5 are based on a growing number of safety-related systems that must be secured with millions of kilometres. All vehicle bus communications and raw data from sensors, cameras, lidar and radar, as well as status data like weather and actual maps must be recorded, authentically. An eight-hour test drive easily produces 4TB of data, if not 20-100TB. The data must then be fed to data centers without causing long pauses in vehicle testing. The enormous amount of data becomes a challenge for measurement equipment in automotive environments. This presentation shows new developments.
founder and principal
Synopsis: Developing automated vehicles that work well 99% of the time is relatively straightforward. But developing automated vehicles that work 99.99% of the time is an order of magnitude more challenging. Recent accidents point to the need for heightened safety and redundancy to cope better with edge cases not normally encountered. In this session, VSI will discuss the latest trends toward improving the safety and performance of automated vehicle solutions.
Kris De Meester
VP sales and business development
Synopsis: Technology choices for automotive lidar have to be based on how and where the lidar units can be integrated in vehicles. This integration has an impact on the lidar technology as well as the vehicle itself. As the need for reliable solid-state lidar systems continues to grow, considerations about sensor placement and ways to achieve reliable detection receive the automotive attention. Although test vehicles often still have externally mounted lidar systems, mass-production vehicles are in need of lidar technology that can be nicely incorporated in the vehicle. This presentation will report the outcome from cooperations between XenomatiX and Tier 1 partners.