Conference Program


Autonomous Vehicle Test & Development Track


Room A Strategies, innovations and requirements for the safe deployment of ADAS and autonomous technologies
09:00 - 12:25

Moderator

Chris Reeves
Head of connected & autonomous vehicles
HORIBA MIRA
UK

09:00

How European legislation paves the way towards driving automation

Maria-Cristina Galassi
Scientific project officer
European Commission
Italy
Connected and Automated driving brings the promise of improved road safety, lower emissions and higher social inclusion. Nevertheless, automated vehicles are very complex systems operating in a very complex environment: regulators need to shape new policies and measures to maximize positive outcomes from driving automation and to avert or mitigate negative ones. With this aim, the European Commission is contributing to the global regulatory effort ongoing at United Nations on L3 highway applications, and is proposing a new EU Implementing Regulation for the type approval of Automated Driving Systems (ADS), enabling the deployment of L4 vehicles on European roads.

09:25

Safety argumentation of autonomously driven vehicles

Ali Nouri
Senior system safety engineer in autonomous driving
Volvo Cars
Sweden
Autonomously driven vehicles (AD) with no human driver are said to offer a fruitful path toward achieving Vision Zero, i.e. no fatal or severe road accidents. But safety argumentation of such a novel, complex and dynamic system is impossible in the context of today’s methods such as those prescribed in ISO 26262. The presentation will focus on the contribution of each of the current standards and guidelines, and Volvo Cars’ safety strategy and concept for having safe AD.

09:50

A coordination framework for R&I and large-scale testing and evaluation, for CCAM in Europe.

Dr Stephane Dreher
Senior manager - connectivity and automation, blockchain
Ertico - ITS Europe
Belgium
The EU funded ARCADE Coordination and Support Action has set up a comprehensive Knowledge Base for CCAM related activities in Europe and beyond. It gathers an overview of R&I projects and demonstration activities, regulations and policies, strategies and action plans, standards as well as methodologies for evaluation and data sharing. The next step will consist in building on experiences and lessons learned from initiatives across Europe to develop a European Framework for Testing on public roads, comprising a Common Evaluation Methodology and a Test Data Exchange Framework. Maintaining and improving mechanisms for exchanging best practices and lessons learned, is essential to ensure a coordinated and harmonized deployment of CCAM in Europe. This presentation will provide an overview of the latest findings and discuss the next steps.

10:15 - 10:45

Break

10:45

The International Alliance for Mobility Testing and Standardization (IAMTS) – a powerful alliance to enable internationally scalable deployment of L4/L5 vehicles

Dr Joachim Taiber
Managing Director International Operations
IAMTS (International Alliance for Mobility Testing and Standardization)
Austria
The presentation will address the key challenges of safety validation of automated and connected vehicles from a methodological and standardization perspective. It will as well deal with the need and setup of an alliance between key standardization institutions, 3rd party technical service providers as well as technology and infrastructure developers, providers and operators. Also, latest outcomes, practical examples as well as further focus areas and topics of the alliance will be presented.

11:10

The Sunrise Project (safety assurance framework for connected, automated mobility systems)

Xavier Sellart Ortega
Manager ADAS and AV development
Idiada Automotive
Spain
Validation methods and testing procedures still lack the appropriate safety assessment criteria needed to build a robust safety case. These must be set and must be valid for the comprehensive scenario parameter space. The potential of encountering different approximations per global region adds an additional level of complexity. Evolving from the achievements obtained in HEADSTART and taking other initiatives as a baseline, it is necessary to move to the next level in the concrete specification and demonstration of a commonly accepted safety assurance framework (SAF) for the safety validation of CCAM systems.

11:35

First cabless freight

Tomas Ohlson
Founding engineer
Einride AB
Sweden
The presentation will outline Einride’s approach to rollout plus the safety issues around putting autonomous freight vehicles on the road.

12:00

A localization stack for L4 autonomous driving

Dr Ralph Grewe
Product manager perception – innovation line driverless
Continental
Germany
In a L4 vehicle, many modules along the AD stack need a precise estimate of the vehicles position or state. Most of the modules need a smooth estimate of relative motion and state but some also require a precise position in a globally fixed coordinate system. To address those requirements, we present a localization stack split into specific fusion systems for relative and absolute localization. Additional to wheel speeds, IMU and GNSS we also use data from radar sensors to improve localization. We show results from real world test drives for both parts of the localization stack across different platforms.

12:25 - 13:55

Lunch

Room A Scenario-based testing and advanced simulation
13:55 - 16:55

Moderator

Dr Sandeep Sovani
GTM head HPC for engineering modeling & simulation
Amazon Web Services
USA

13:55

The ever-developing testing landscape for ADAS and AD

Benjamin Engel
Global technology manager
ASAM eV
Germany
Both virtual and scenario-based testing continue to gain traction as primary testing techniques for multiple goals in the industry, including for homologation and safety argumentation. What implications do these ‘new kids on the block’ have for other, more established testing methodologies and workflows? What does the industry need to take into account when including these techniques in a full-featured testing campaign? What does this developing landscape require from technical services and standardization organizations? This presentation will give insight into ASAM’s latest study group on test specifications, released in Q1 2022. It aims to start answering these questions and more, to kick-start further discussions and activities around testing and validation for AD. It will go over some of the next steps the industry and standardization organizations need to take to provide a fully-fledged collaborative testing approach for ADAS and AD.

14:20

Enabling virtual test & validation - creating a proof of validation by empowering decentralized open data markets within the GaiaX research project

Carlo van Driesten
Systems architect for virtual test and validation
BMW Group
Germany
Achieving a virtually enhanced homologation process through the usage of driving simulators relies on the quality of the data. Integrating elements such as HD Maps, driver models, scenarios, or sensor models, in various systems throughout deep supply chains demand a common understanding brought by standards, like e.g., OpenDRIVE from the ASAM e.V. OpenX. But what makes a map “HD” and how can we validate the quality of such a claim, e.g. in the context of ALKS regulation UN R157? High quality data in standardized formats can also be used as a corner stone of simulation tool certification. How can we access and increase the availability of this data? BMW explores the possible creation of an open and decentralized data ecosystem (ODDE) within the European research family GaiaX, using DLT technologies for self-sovereign identities leveraging zero-knowledge proofs for selective privacy in open networks.

14:45

Using ASAM OpenSCENARIO 2.0 and the scenario-based coverage-driven verification approach to efficiently tackle the infinite scenario space

Gil Amid
Chief regulatory affairs office
Foretellix
Israel
The presentation describes how the recently published OpenSCENARIO 2.0 standard 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 standards enable a step function in ensuring the correctness and safety of automated driving systems.

15:10 - 15:40

Break

15:40

Traffic simulation - modeling high-level driving decisions from naturalistic driving-data

Dr Christian Contreras Campana
Expert research scientist
Automotive Artificial Intelligence (AAI) GmbH
Germany
Traditional driving stacks have inherent shortcomings in generating trajectories that are both feasible and naturalistic at high and low abstraction levels. In this presentation we propose a new model for simulated traffic agents that is able to mimic high-level decisions from naturalistic traffic data and can employ any trajectory planning method. The agent model focuses on human-like behavior, decision-making, and driving performance based on a supervised machine learning approach. We will show that even though our approach is based on modeling nanoscopic traffic-agent characteristics, realistic traffic behavior can be achieved both at the microscopic and macroscopic levels.

16:05 - 16:55

Panel discussion - Next frontiers in ADAS and autonomous vehicles simulation

Simulation is an indispensable tool in the development of ADAS and Autonomous Vehicles. Simulation tools and processes must keep pace with the rapid advancements happening in ADAS and AV technology stacks. This panel will share unique insights into the current state of ADAS and AV simulation technology and the next frontiers that the field faces. We will hear expert perspectives from an automotive OEM, a Tier-1 supplier, an AV simulation company, and a prominent academic thought leader. The panel will dive deep into new capabilities, process improvements, and innovations in simulation technology that will address upcoming challenges in all aspects of the automated driving stack including sensing, perception, planning, and controls.
Varun Mittal
General manager
Applied Intuition
USA
Carlo van Driesten
Systems architect for virtual test and validation
BMW Group
Germany
Dr Richard Rapp
CDV simulation manager
Valeo
Germany
Dr Kmeid Saad
Senior Manager Application Engineering
Ansys
Germany
Moderator:
Sandeep Sovani, GTM head HPC for engineering modeling & simulation, Amazon Web Services

Room A Real-world testing and deployment
09:00 - 12:50

Moderator

Christian Thiele
Director global ground vehicle standards
SAE International
USA

09:00

Bringing autonomous vehicles to the roads of Scandinavia

Christian Bering Pedersen
CEO
Holo
Denmark
There is plenty of hype around autonomous vehicles, but what is the current situation in terms of practical implementation? Christian will present the current status of Holo’s work on bringing autonomous shuttles to the roads of Norway, Sweden and Denmark. Challenges include getting the right approvals from authorities, finding the right routes, operating in adverse weather conditions, training and managing a crew of safety drivers (still required by law), collecting and analyzing operational data and many more. Reality doesn’t look like the hype – but the current reality of AVs is still pretty exciting.

09:25

Proving ground requirements for automated vehicle testing

Nils Katzorke
Project coordinator
Mercedes-Benz AG
Germany
Due to the higher number of components and versions in automated vehicles (AVs), the number of required tests increases as well. While digital methods like model-in-the-loop and software-in-the-loop gain ground, physical validation remains mandatory. The presentation will provide an overview of major future milestones in the development of automated vehicles and their effects on testing requirements. Furthermore, it will highlight future changes in requirements for proving grounds and approaches to identify these. As an example, some results and measures for the new Mercedes-Benz proving ground in Immendingen will be shown.

09:50

Co-creating the future of road transport: the JRC Living Lab for Future Mobility Solutions

Dr Biagio Ciuffo
JRC project portfolio leader - smart mobility
European Commission
Italy
The Living Lab is an open innovation ecosystem taking place through a public-private-people partnership. It aims to develop new products, services, businesses or technologies through active user involvement and co-creation at the early stages of the innovation process by creating, testing and validating innovations in real-life settings. The Living Lab allows the JRC to support innovative start-ups while gaining useful insights to support the EU policy development process in the transport field. In 2019, the JRC made a pilot call to engage EU entities in proposing possible ways to use its Ispra site, to co-create future mobility solutions. This presentation will provide an overview of the projects carried so far, added value generated and future opportunities.

10:15 - 10:45

Break

10:45

Guidelines for Testing Automated Driving Systems in Canada Version 2.0

John Conway
Policy analyst
Transport Canada
Canada
To achieve the many positive outcomes promised by automated vehicle (AV) technology, it is essential to safely conduct testing in real-world environments. In 2021, Transport Canada published a revised guidance document titled Guidelines for Testing Automated Driving Systems in Canada Version 2.0. This updated guidance will help to ensure that AV trials continue to be conducted safely based on lessons learned from testing activities within Canada and around the world. This presentation will provide an overview of the safety best practices covered in the updated guidance, including safety considerations associated with remote support as well as low-speed automated shuttles.

11:10

Deployment of automated drive functions in vehicles

Plato Pathrose
CTO, Vantix Inc. and chief engineering group ADAS/AD
Vinfast LLC
Germany
The presentation covers the testing and deployment of automated driving functionalities in vehicles. This includes the practical approaches that were followed when deploying automated driving functionalities, how the maturity of various features was measured and how the features were evaluated to be safe enough. Rather than being theoretical, the case studies and presentation cover the deployment of vehicles at various geographical locations and explain how the quality and safety aspects were organized. We will also outline the features of the Vinfast vehicle and how their quality and safety are guaranteed.

11:35

Challenges in real-world regulatory type testing of AVs in Singapore

Paul Hibbard
Senior research engineer
CETRAN - Centre of Excellence for Testing & Research of Autonomous Vehicles – NTU University Singapore
Singapore
CETRAN (Centre of Excellence for Testing & Research of Autonomous Vehicles – NTU) has been working with the regulatory authorities in Singapore since 2017 to create a framework for and to conduct testing of autonomous vehicles (AVs) to enable R&D deployment of AVs in Singapore. The presentation will cover the approach that has been developed and some of the challenges experienced through real-world experience of conducting physical tests of AVs. This will include physical scenario-based testing at the CETRAN proving ground test track and also deployment testing on the public roads in Singapore.

12:00

ADAS testing: using advanced choreographed robotic vehicles

Jaren Devey
Automotive development lead
Autonomous Solutions, Inc.
USA
ASI has developed autonomous systems capable of choreographed testing of ADAS technology. Driven by robotic systems, these autonomous vehicles are capable of tighter, more accurate positioning than manned vehicles can safely perform. This technology utilizes vehicle-to-vehicle and vehicle-to-server communications with advanced control algorithms ensuring autonomous vehicles will be posed at predefined positions and velocities at given times. This fully exercises the responses of the entire ADAS system required for autonomous driving. When triggered, the ASI autonomous system releases control to the ADAS system while continually monitoring safety system, ready to regain control instantly if conditions are not nominal.

12:25

Waymo's approach to ensuring safe deployment of fully autonomous driving technology

Dr Trent Victor
Director of safety research and best practices
Waymo
USA
In October 2020, Waymo opened its fully autonomous ride-hailing service, Waymo One, to the general public in Metro Phoenix, Arizona. Right now, members of the public are hailing vehicles with no human driver controlling the car – either in the vehicle or remotely – to help them get to where they’re going as part of their everyday lives. In the past 12 months, Waymo has expanded its operations to new territories, including San Francisco and Downtown Phoenix. Since its start in 2009 as the Google Self-Driving Car Project, Waymo has spent years refining a comprehensive set of robust methodologies that assess safety across its technology, operations and team culture, and ultimately guide the deployment and safe operations of the Waymo Driver. In this talk, Dr Trent Victor, Waymo's director of safety research and best practices will discuss the safety framework that guides Waymo's progress and is the first of its kind in the AV industry.

12:50 - 14:05

Lunch

Room A Innovations in simulation, validation and integrated physical testing
14:05 - 17:05

Moderator

Sytze Kalisvaart
Senior project manager StreetWise
TNO
Netherlands

14:05

An integrated, system-level approach for the testing and validation of vehicle and infrastructure-based cooperative driving automation (CDA) technologies

Dr Mircea Gradu
Professor mechanical and aerospace engineering
University of California
USA
Dr Blake Lane
Senior scientist, advanced power and energy program
University of California
USA
Cooperative driving automation (CDA) aims to improve vehicle and general traffic safety and energy efficiency, and to facilitate road operations. CDA greatly benefits from high levels of vehicle automation and from sophisticated ADAS technologies. Utilizing the extensive analytical and physical testing and simulation capabilities of the Horiba Institute for Mobility and Connectivity at the UCI School of Engineering, this project replicates in a lab environment the sensing and perception algorithms employed on automated vehicles and on the infrastructure of a public road network platform, to facilitate the early assessment of their effectiveness and to define the best strategies for implementation.

14:30

Driving simulation for the evaluation of car sickness in autonomous vehicles

Dr Andras Kemeny
Expert leader immersive simulation
Renault
France
If future self-driving systems, with complete delegation of driving, may demand redesigning vehicle interiors and corresponding user interfaces, most personal autonomous vehicles will, for some time, keep more traditional vehicle architectures. In self-driving cars, a significant part of users may then experience car sickness, and this safety hazard may become a crucial social issue, during driving delegation. Driver in the loop (DIL) simulation, already used to evaluate autonomous vehicle acceptation, could then be used for car sickness experimentation. Various motion based DIL simulation experiments are already carried out, however the possibility to differentiate simulator and car sickness effects determines the validity of such tools in function of the studied uses cases.

14:55

Vehicle, sensors, traffic and scenario: how to trust simulation models?

Thomas Nguyen That
Software products director
AVSimulation
France
While simulation plays a strategic role in the development and validation of ADAS and AD systems, being more and more used all along the V-cycle, calibration and correlation of models is crucial to being able to trust the simulation results. This presentation will explain the methodologies we have developed to build and validate our models including virtual roads, vehicles, sensors, traffic and scenarios: from field tests on testing tracks or real driving, through to characterization in the lab.

15:20 - 15:50

Break

15:50

Road to lab: the validation of ADAS/AD functions in automotive

Henrik Liebau
R&D manager, autonomous drive emulation test automotive and energy solutions
Keysight Technologies
Germany
The automotive industry is facing tremendous change and multiple technology disruptions. Unlike the mechanical age, testing can no longer be performed on the road alone. Despite the progress in scenario and vehicle dynamics modeling to stimulate the visual sensors (cameras), radar, lidar and V2X emulation are still far behind. Besides the technical limitations of the various sensor emulation implementations, their integration into a complete and comprehensive environment is extremely complex. Attend this session to learn about Keysight’s approach to autonomous drive emulation.

16:15

Safety Pool Scenario Database: enabling an ODD-based testing framework for automated driving systems

Dr Siddartha Khastgir
Head of verification & validation, Intelligent Vehicles
WMG, University of Warwick
UK
This presentation describes how the Safety Pool Scenario Database (the world’s largest public scenario database for automated driving systems) can be used as part of an ODD-based (operational design domain) testing and certification framework. Catering to a variety of use cases in diverse ODDs, the Safety Pool Scenario Database will enable test engineers, developers, regulators and policy makers to seamlessly link ODD definition (which will be a prerequisite for AV regulation) with test scenarios and corresponding metrics for “safe enough”. Furthermore, being aligned to various international standards, the database enables interoperability for virtual verification and validation and real-world testing.

16:40

Calibration of car traffic bringing RDE tracks close to reality

Jochen Lohmiller
Principal product manager
PTV Group
Germany
Virtual simulation needs to replicate real-world situations as well as possible. When performing simulations with traffic, often only the individual behavior of the surrounding traffic is of importance and general traffic patterns are neglected. That is, for example, realistic paths of vehicles that determine the weaving/merging rates but also realistic speed-flow relationships. In this presentation, we introduce how to calibrate traffic simulations using origin-destination matrices based on GPS data, video footage and other sources to achieve a realistic representation of traffic flow in your virtual tests.

Room A Validation, verification, data management and in-the-loop testing
09:00 - 16:55

Moderator

Holger Banzhaf
Managing director
DeepScenario GmbH
Germany

09:00

Comparability and exchangeability of scenario databases

Sytze Kalisvaart
Senior project manager StreetWise
TNO
Netherlands
Scenarios have become central to AV safety assessment to describe situations an AV will meet. TNO collects scenario statistics from driving data. These help to quantify a positive risk balance – is the new generation safer? Now that driving data collection is established at OEMs and tiers, the next step is comparability and exchangeability of such scenarios and statistics. TNO compares databases at the level of parameter distributions to identify biases or blind spots. This helps to direct data collection and completeness for legal due diligence. A consortium of industrial partners has started to exchange scenario statistics, coordinated by TNO.

09:25

AV beyond the hype: how to prepare for mass deployment

Robin Van der Made
Director product management
Siemens Industry Software
Netherlands
The concept of AV has been proven in many pilots during the last decade. The first companies are offering commercial AV services at a limited scale. Now, the industry needs to prepare for scaling up. These vehicles will serve new locations with new driving environments, such as more variable weather and road conditions, diverse driving cultures and further challenges. How do we deal efficiently with this increase, especially in the light of scarce engineering resources around the globe? This presentation will show how to grow the expansion of AV efficiently without growing the underlying development and validation resources at the same pace.

09:50

Generating synthetic data for neural network training

Daniel Tosoki
Product & operations manager of aiSim
aiMotive
Hungary
Real-world data is vital for training detection algorithms. However, gathering data on edge cases for training purposes is not always easy – sometimes it isn’t even possible. Furthermore, no matter how much training data is collected, it will always be unbalanced, and there are dangerous and complicated scenarios (for example, close car data) to be collected, which should and can be replaced by simulation. We will show on real examples and measurements how synthetic data generated by aiSim, the world's first ASIL-D certified validation tool, helped overcome these difficulties and improved the performance of neural network-based perception.

10:15 - 10:45

Break

10:45

HIL simulations reach a higher dimension of realism

Gordan Galic
Technical marketing director
Xylon
Croatia
Matt Daley
Operations director
rFpro
UK
Xylon and rFpro teamed up to explore methods of connecting hardware ECUs to simulation software. In this presentation, we will describe how we immerse a real-world surround-view parking assistance ECU in a simulated world seen by four fully modeled virtual HD video cameras placed on the vehicle model driving along virtual roads. This ADAS ECU was selected based on its highly visual nature and required transfers of large amounts of video data. We will outline the challenges and solutions for quick and precise translations between physical and virtual domains that enable this full-speed, closed-loop HIL simulation setup.

11:10

Do you have the right data for AV development?

André Rolfsmeier
Director - strategic product management
dSPACE GmbH
Germany
Petabytes of data are recorded during AV development and the costs for managing the data are immense. In this talk, we will explain how to explore, tag and manage huge amounts of data efficiently. By means of AI-based detectors, you will gain a deep understanding of what data you already have and what data is still needed to develop and test safety-critical perception and autonomous driving systems. Using our own data acquisition vehicle, we collect 30+ TBs of data per test drive. Our experience in automatically detecting defined traffic scenarios in real time during driving or in post-processing will be shared.

11:35

Virtual assessment framework and ontology-based test case generation methods

Dr Jianbo Tao
Senior lead engineer ADAS/AD XiL simulation
AVL List GmbH
Austria
Dr Heiko Scharke
Senior global product manager for ADAS/AD solutions
AVL List GmbH
Austria
A comprehensive and reliable testing process for automated and autonomous driving functions must follow a reasonable path from parameter selection and test scenario generation to test execution and final assessment of the achieved test coverage. In the presentation we will provide an overview of our ontology-based testing approach that is based on five main pillars: the scenario sources, the ontology model setup, the automated test scenario generation and planning, the test execution and evaluation framework and the test management system. The test management system ensures continuous monitoring of the testing.

12:00

1+1=3: hybrid HIL systems for cost-effective ADAS validation

Johann Führmann
Head of business development
b-plus automotive GmbH
Germany
Due to the UNECE-VMAD working party, hybrid HIL systems can now be used to mix real-world driving with simulated driving. At first glance, a lot of validation expenses can be saved by this, but many of the pitfalls will stay the same. Setting up HIL clusters can be time-consuming and expensive in addition to requiring a lot of space in the data centers. Reusing existing SIL hardware can decrease the initial expenses significantly. A thoughtful choice of farm location, test mixture and maintenance strategy will ensure cost control right from the start.

12:25 - 13:55

Lunch

Moderator

Dr Clara Marina Martinez
Engineer - ADAS virtual development
Porsche Engineering Services
Germany

13:55

We keep you safe on the road (dynamic risk management)

Dr Thomas Freudenmann
Co-CEO/co-founder
EDI GmbH - Engineering Data Intelligence
Germany
The RELAI – Risk Estimation with a Learning AI – project's approach is to analyze real-world driving scenarios and the driving behavior necessary for safe driving, which takes into account not only the official traffic rules but also the traffic behavior of other road users and generates test scenarios from this data. The made system in the RELAI project is backed up by the dynamic risk management (DRM) system, which enables autonomous driving systems to dynamically manage various risks in road traffic as experienced human drivers do. This allows the autonomous vehicle to make important decisions based on the driving context.

14:20

Real-time over-the-air automotive radar hardware in the loop to test autonomous vehicles

Dr Kasra Haghighi
CEO
Uniquesec
Sweden
Fahimeh Rafieinia
CTO
Uniquesec
Sweden
Autonomous driving will revolutionize the future of mobility. The main pillar of autonomous driving, as well as advanced driver assistance safety systems (ADAS), is sensing. Among all sensors, radars are the most reliable and versatile ones, providing environment perception. Test and validation of radars and radar-based safety functions is a necessity that requires an advanced OTA radar HIL setup where rich and dynamic traffic scenarios can be emulated. Certain real-time requirements need to be established between environment simulator and radar-target simulator. This presentation will demonstrate a full OTA HIL setup enabling emulation of more than 200 moving targets for the radar under test.

14:45

Left-shifting the validation of autonomous driving using simulation

Jihas Khan
Senior specialist
Tata Elxsi
India
Simulation is an ideal choice for the left-shifting of AD validation, ensuring quality, cost and time savings. Software-in-the-loop testing, synthetic scenario simulation, virtual ECU development and hardware-in-the-loop testing are various types of virtual validation of AD, which will be explained with use cases in the presentation. However, ensuring process compliance and uniform data exchange is of paramount importance for these virtual simulation scenarios. The presentation will also touch on how the ASAM open framework can play a vital role in ensuring the uniformity of test and simulation frameworks. The presenter will show case-relevant live work samples.

15:10 - 15:40

Break

15:40

How can we better streamline test and validation processes?

Simon Tiedemann
Senior product manager - EB Assist Test Lab
Elektrobit
Germany
Bastian Speth
Senior business manager
Elektrobit
Germany
We face many challenges in how we record and consume data. We’re finding inefficiencies in the process with incomplete or unknown data. Elektrobit has proven expertise in data collection and processing, with hardware and software used in millions of vehicles today. Are storage costs increasing with each project? While conducting test drives all over the world, Elektrobit has developed innovative solutions that have saved customers significant time and money. This presentation will share how a new data orchestration platform can help collaboration and data socialization, connecting data scientists, managers, engineers, testers and algorithm developers to achieve common goals.

16:05

Sensor fusion simulation using 3D imaging radar and lidar

Koki Taniyama
Engineer
OTSL Inc
Japan
Recently, 3D imaging radar performance has been increasing, and this point cloud data will catch up with the lidar point cloud in the near future. OTSL has developed the new MIMO Radar Simulator (COSMOSIM-AMMWR2) for supporting 3D imaging and point cloud, and has implemented various Tier-1 dedicated radar models for more accurate simulation. The presentation will showcase show genuine MIMO radar simulation in real time.

16:30

A distributed simulation for autonomous vehicle test and development

Amir Soltani
Founder and managing director
KAN Engineering Ltd
UK
The critical role of simulation in the successful development, test, verification and validation of connected and autonomous vehicles is now being widely acknowledged. Still, no single simulation tool is available that can cover the entire development V-cycle. Under Zenzic's interoperability projects, KAN Engineering has delivered a novel solution for an integrated end-to-end distributed interoperable simulation architecture across the UK. The developed technical solution enables seamless deployment between various stages of development from MIL, SIL, HIL, DIL, VIL, test tracks and real-world roads, to carry out similar tests in different simulation contexts and geographical locations, all from a single place.

Room B Connectivity, mapping and positioning
09:00 - 11:35

Moderator

Benjamin Engel
Global technology manager
ASAM eV
Germany

09:00

Mapping change: new approaches to map maintenance for AD applications

Nicola Consolati
Head of OEM delivery
Woven Planet
Japan
Map maintenance - the updating of maps to account for changes in the road - has long been one of the most challenging aspects of automotive mapping, especially for automated driving applications. This complexity is becoming ever greater as automated driving technology proliferates and accurate, reliable automotive-grade mapping is needed at continental scale. In this presentation, Nicola Consolati will help size this ‘change’ problem and discuss new, multimodal approaches to managing change, using examples from recent work. The presentation will pay particular attention to the challenges presented by urban mapping and the technical solutions needed to map cities at scale.

09:25

V2X technology developments, testing requirements and certifications

Thomas Jaeger
Senior vice president
Dekra SE
Germany
The presentation will cover the latest V2X technology developments (ITS-G5, DSRC-wave, cellular), the latest regulatory and certification situation (EU, US and others), testing requirements for connected and automated driving (regulatory, conformance, use cases, interoperability), the most recent developments in regional and global interest groups featuring certifications, and challenges and the outlook for the future.

09:50

Distributed Mobility Cloud: From safety to automated driving support

Uwe Puetzschler
Head of Car2X
Nokia Mobile Networks
Germany
Starting with sharing of safety related traffic information the automotive ecosystem will have to evolve and move to more advanced use cases supporting automated driving and VRU protection. Edge computing will play an important role for the deployment of those advanced use cases. Experience gained during test deployments and the role of data for sustainable cooperation and business models will be presented and discussed.

10:15 - 10:45

Break

10:45

Assessment of GNSS-based equipment in the context of AEBS-UNECE R152

Ernst Phillip Mrohs
Head of laboratory
NavCert GmbH
Germany
With regulations in the areas of the pan-European-eCall, EETS and digital tachograph, obligatory requirements for GNSS-based in-vehicle systems are mandated. Similar is specified in the field of driving assistance, e.g. UNECE-R152. Conformity is verified in accredited laboratories. In each laboratory, the calibration of measuring instruments is essential for testing reliability. Since GNSS observations of such instruments are not directly traceable to SI units, an assessment, validation and qualification must be performed. For this, NavCert developed a procedure based on existing standards. The core is performance verification. An exemplary assessment of the PVT solution is presented here.

11:10

Vehicle platooning with dynamic performance optimisation

Tim Edwards
Chief engineer - connected autonomous vehicles
Horiba MIRA
UK
Vehicle platooning presents an opportunity to leverage automation and connectivity in combination to operate safely and efficiently in highway driving. The PLATONIC project, a collaboration between HORIBA MIRA and GMV NSL, funded under the ESA Navigation Innovation and Support Program (NAVISP) stimulating innovation in the PNT domain, addresses advances in cooperative positioning and integrity algorithms for platooning applications. This presentation will describe the project approach with a focus on the design and validation challenges, and how real-time integrity monitoring can be used for dynamic optimisation of vehicle speeds and following distances.

Room B Best practices, standards and regulations for the deployment of safe ADAS and autonomous driving technologies
11:35 - 16:55

Moderator

Benjamin Engel
Global technology manager
ASAM eV
Germany

11:35

The latest update on connected and autonomous driving regulations: global overview and implications

Alexander Schwab
Officially recognized expert for motor vehicle traffic
TÜV SÜD
Germany
With UNECE R157, an international framework was created in 2020 for the approval of an automated lane-keeping system (according to Level 3). In 2021, Germany introduced the possibility of permitting Level 4 vehicles on public roads by adapting the Road Traffic Act (StVG). The corresponding draft directive is in the final adoption process. This session will give an overview of the current regulations worldwide. We will present the basic ingredients of effective approval, considering virtual and physical tests, SOTIF and cybersecurity, using the example of the German Level 4 Regulation, highlighting the challenges and possible approaches to solving them.

12:00

Reference measurement system for development and validation of ADAS functionalities

Steffen Metzner
Product Manager ADAS Reference Measurement Systems
AVL List GmbH
Austria
Despite all efforts to carry out tests of vehicle automation functions using simulation technology, the major part of tests and the validation is still necessary to be executed directly in-vehicle. For an efficient development and verification of sensor and perception systems a highly accurate ground truth reference sensor device is necessary. Without such systems, troubleshooting and objective evaluation of vehicle behaviour becomes very expensive and often even not possible. The solution discussed in this presentation aims at supporting the development of vehicle automation functionalities during the entire development process from sensor selection to function validation.

12:25 - 13:55

Lunch

Moderator

Adrian Bertl
Teamleader product marketing
b-plus technologies GmbH
Germany

13:55

A new measure of AV Safety: Are existing measures good enough?

Dr Erik Antonsson
Founder and CTO
Streetscope Inc
USA
A new quantitative measure of safe operations in traffic, based on continuous measurement of the positions and movements (kinematics) of traffic objects has been developed and is being deployed. This new approach provides a dense, continuous, inexpensive, assumption-free way to measure, assess, and compare driving behavior, e.g., different builds of an AV developer’s software stack, or different configurations of sensors, and overcomes important problems with existing measures such as TTC, ISM, RSS, etc. A quantitative measure of this type is essential to measuring engineering progress and to providing an independent, quantitative basis for safety standards and public acceptance of AVs.

14:20

Tactical safety and its implications for autonomous vehicle testing

Hans-Peter Schöner
CEO
Insight from Outside Consulting
Germany
Situation awareness allows good human drivers to judge when to be particularly cautious, on top of following standard driving rules, deliberately undertaking additional actions such as reducing driving speed and increasing distances to other traffic participants. This behavior, denoted as tactical safety, allows the avoidance of constellations that could not be controlled safely, or that could hinder fluid traffic movement. When applying tactical safety strategies to autonomous vehicles, communication between the vehicle and other traffic participants and/or with the infrastructure is of particular importance. Finally, to enable the technical implementation of tactical safety strategies, the development of quantitative measures for the danger of specific situations is a key issue. This presentation will elaborate on tactical safety and its implications for autonomous vehicle testing. Co-Author Dr. Jacobo Antona-Makoshi, Japan Automobile Research Institute (JARI), SAKURA project

14:45

Level 2-3-4-5 overview of regulation/Euro NCAP status and testing challenges

Alain Piperno
Senior expert - autonomous vehicles testing and homologation
UTAC
France
UTAC is a worldwide reference in testing and type approval, an accredited lab for Euro NCAP tests and for French type approval tests. This presentation will review today’s regulation and testing capacity proposals and new challenges. It will cover Level 2-3-4-5 regulations status for M1 vehicles, shuttles, buses, delivery robots, SW/OTA, cybersecurity, AI and virtual type approval. We will also discuss needs/solutions for testing and validation, including physical, new protocols, new targets, rain and fog testing, digital and VIL testing, AI, V2X and cybersecurity.

15:10 - 15:40

Break

15:40

The law on autonomous driving in Germany – last-mile issue

Jörg Kahler
Attorney / partner
GSK Stockmann
Germany
With the newly enacted Autonomous Driving Act, Germany created the world’s first legal framework for the approval of SAE Level 4 vehicles out on public roads in specified operating areas. The legal certainty now provided is intended to bolster trust in and acceptance of these new technologies. The presentation will introduce the new law and its particular importance for the last-mile issue. It will also outline the remaining challenges regarding the issues of responsibility, liability and data privacy, and how to deal with them.

16:05

Challenges of testing the connected automated/autonomous vehicle – a holistic approach

Dr Alain Vouffo
Senior product manager automotive
Spirent Communications plc
Germany
The connected automated vehicle (CAV) is a system of systems combining a wide variety of wired and wireless communication technologies with different types of sensors to provide advanced applications for enhancing driving comfort and safety for vehicle owners and other road traffic participants. This presentation will discuss the key challenges facing the automotive industry in testing the CAV, along with some potential methods and tools for a holistic approach to addressing those challenges.

16:30

AD management system to fulfill technology and legal compliance

Lucas Bublitz
Senior consultant autonomous mobility / product and process compliance and PhD student for autonomous driving
P3 Automotive GmbH | University of Aveiro
Germany
This presentation will introduce a holistic AD management system (ADMS) framework that defines the safety, security and quality frame in terms of processes, methods, tools and organization through the whole digital lifecycle of AD systems. The framework meets the requirements of the safety and security norms like SOTIF, functional safety, cybersecurity and Automotive SPICE and ensures advanced risk and threat management, which will be continuously adapted, and change/configuration management, which requires fast incident reaction and function validation. An ADMS addresses the challenges of operations of SAE L3 and L4 applications considering the whole development phase, and sets the organizational structure.

Autonomous Vehicle AI, Software and Sensor Fusion Track


Room B Morning Day 1
09:00 - 12:25

Moderator

Abhash Das
Senior safety staff
UL International Germany GmbH
Germany

09:00

Advances and perspectives on AI decision making for CAVs

Joshué Pérez
Research leader
Tecnalia
Spain
Connected and automated driving has increased its functionalities in recent years. Most of the applications receive onboard sensor and communication inputs from the infrastructure and other vehicles. Some decision and control techniques have been implemented for complex environments; however, most of them spend a lot of time in the execution. This presentation describes the techniques used by different research teams and their contributions to motion planning. Moreover, an approach based on a testing methodology for validation of path planning and control algorithms for current and future automated vehicles is presented.

09:25

Leveraging AI in modern sensor fusion applications

Dr Árpád Takács
Sensor fusion group leader - autonomous mobility business area
Continental AI Development Center
Hungary
As AI-based sensor data processing becomes a standard in autonomous driving applications, modern sensor fusion approaches allow the comprehension of complex environments. This opens space for the simultaneous handling of static and dynamic obstacles, ambiguous situations, and a more general solution to edge cases. We will discuss some of these challenging scenarios we come across during our daily work at Continental, and will provide an insight to how we leverage AI-based methods to tackle them in real-world applications.

09:50

AI-based scene prediction for automated driving

Dr Anne Stockem-Novo
Professor
University of Applied Sciences Ruhr West
Germany
Safe handling of automated vehicles requires anticipation of the traffic scene several seconds ahead. In the publicly-funded KISSaF project, a measurement vehicle is equipped for taking data. Within the project, the gained data is used for the development of a robust scene prediction model and testing with standard metrics as well as functional KPIs with a scenario-based test approach under realistic conditions. State-of-the-art deep learning techniques are used to tackle all kinds of scenarios, including highway and rural as well as urban scenarios.

10:15 - 10:45

Break

10:45

Implementing data-driven development for autonomous vehicle development

Martin Eichhorn
Strategist - robotic drive
DXC Technology
Germany
Perminder Singh
Global tech lead - automotive
Amazon
USA
The research and development of ADAS/AD functions is changing the way automotive engineering teams work. Fundamentally larger amounts of data are required for the development, validation and verification of such functions. DXC and AWS have established a global alliance as preferred partners to support autonomous driving development more effectively. This session will give concrete examples of how development workflows can be automated based on large volumes of data, using the combined capabilities of DXC and AWS to build processes, methods and tools for integrated processes with end-to-end automation that can be scaled to a true data-driven development platform.

11:10

Developing and testing autonomous vehicles at scale

Frank Kraemer
IBM systems architect
IBM
Germany
Developing, training and testing AI for autonomous driving is an extremely complex and costly effort. Modern IT infrastructure accelerates end-to-end AI platform development for AVs, encompassing the datacenter hardware, software and workflows needed to develop and validate self-driving technology from data collection through validation. No single company can solve these large challenges alone. It takes open collaboration and partnerships to turn these big problems into solutions.

11:35

Calibrating ADAS and AV vehicles at scale

Mohammad Musa
Founder and CEO
Deepen AI
USA
Maintaining accurate sensor calibration is key to all ADAS and AV systems. Now that most new cars will have ADAS features in them, undertaking multi-sensor calibration at scale is becoming a real bottleneck. The presentation will discuss the multi-sensor calibration lifecycle, all types of calibrations required and best practices for conducting multi-sensor calibration at scale.

12:00

Managing AI models & the ADAS/AD toolchain

Gabriel Sallah
EMEA HPC and big data architect, autonomous driving platform solutions
Microsoft
United Arab Emirates
The ADAS/AD toolchain has proven to be a very complex environment to effectively manage and scale . The pace of innovation around sensors, in car technology requires a flexible, open and highly resilient IT infrastructure and tooling to supports. This talk will describe Microsoft Architects' ADAS/AD test and validation platform on the cloud. There will be special emphasis on how Microsoft map in Car SOC technology to hardware acceleration devices on the Azure cloud. It will look at how to train and manage AI models and running ReSim/Replay and Simulation at scale using traditional High Performance Computing and data techniques. The talk will close with the lessons learned in ADAS in Simulation and how it currently being used as the foundations for the Metaverse.

12:25 - 13:55

Lunch

Room B Afternoon Day 1
13:55 - 16:55

Moderator

Mohammad Musa
Founder and CEO
Deepen AI
USA

13:55

Co-opetition: rapid innovation through collaboration

Dan Cauchy
Executive director, Automotive Grade Linux
The Linux Foundation
USA
Auto makers are now in the tech business with the advent of connected and autonomous vehicles, but lengthy production cycles hinder them from innovating at the speed of the tech industry. While some auto makers are forging partnerships or making acquisitions, others are embracing co-opetition through open-source software like Automotive Grade Linux (AGL) to drive rapid innovation. AGL unites the automotive and tech industries around shared software for all in-vehicle applications from infotainment to autonomous driving. Sharing a single software platform across the industry decreases development times so OEMs and suppliers can focus on rapid innovation and bringing products to market faster. This presentation will discuss the rise of open source in automotive and why companies that are traditionally competitors are working together to develop shared technologies. It will also provide an overview of AGL including key milestones, the project roadmap and how to get involved.

14:20

Progress on the AUTOSAR Adaptive Platform for Intelligent Vehicles

Rinat Asmus
Steering committee member AUTOSAR development partnership and AUTOSAR chairperson
BMW Group
Germany
AUTOSAR develops completely new approaches to cope with the challenging market trends in the automotive industry, like internet access in cars, highly automated driving and vehicle-to-vehicle communication. The result of these activities during the last four years is an intelligent and flexible infrastructure software that is therefore named AUTOSAR Adaptive Platform. The platform runs on high-end computing hardware and supports parallel processing on many core systems and GPUs. Consequently, it can also be used to support high-bandwidth communication and is able to host AI applications.

14:45

Steps for safe road release of ADAS/AV functions

Abhash Das
Senior safety staff
UL International Germany GmbH
Germany
A growing number of technology companies release autonomous driving functions on an ad-hoc basis. This brings confusion to end users. Lack of established standards for these mostly prototype software functions brings additional risk to society. New manufacturers of such safety-critical functions perform road releases with a lack of transparency and oversight. Some of the established OEMs and Tier 1s have ways to roll out such functions as prototype releases thanks to their years of experience, which the new players in the automotive field do not have. This presentation will discuss the real challenges that manufacturers face and steps to ensure safe road release.

15:10 - 15:40

Break

15:40

The automotive industry is shifting toward a data-centric, network-agnostic architecture

Pedro López Estepa
Market development director, automotive
Real-Time Innovations
Spain
As OEMs race to develop connected and autonomous vehicles, software-defined architecture takes center stage. Yet the pervasive in-vehicle connectivity poses new risks, particularly around vehicle security and safety. Off-vehicle, the industry now demands reliable, secure and non-intrusive communication for real-time teleoperation and data analysis at high data rates and fidelity. In this session, RTI will present the capabilities and challenges of in-vehicle communication as well as how to leverage the latest telecommunication infrastructure. RTI will also demonstrate how, thanks to its unique network technology, Connext Drive represents the best communication solution over mobile and wide area networks.

16:05

Building a digital law pipeline for machine-executable legislation

Capm Petersen
Chief innovation officer
Verses Labs
USA
Stephen Lernout
Head of strategy & innovation
NALANTIS
Belgium
Nalantis and Verses are partners in FlyingForward 2020, a consortium funded by the European Commission and working on a new urban air mobility ecosystem and aligning it within the spatial data infrastructure of cities. We are applying natural language understanding AI to automatically deconstruct human written legislation into meaningful semantic code and integrate this code with an open geospatial digital infrastructure based on spatial web standards. Currently, a vast amount of manpower and labeled training data for machine learning purposes is still needed to code this. Our methodology creates plug-and-play machine-executable algorithms that can be applied to the autonomous space.

16:30

Autonomous and AI in the evolving regulatory landscape

Dr Alexander Duisberg
Partner
Bird & Bird LLP
Germany
Germany has laid out the regulatory landscape for autonomous drive at the forefront of the market. At the same time, the EU AI Act is coming through in 2022, the Product Liability Directive is up for discussion, and the Data Act is pushing for new rules on data access and data sharing. How do these provisions fit and work together? What are the implications for liability of OEMs and operators of ‘high-risk systems’? How does data sharing work legally to train AI algorithms? The presentation discusses various aspects of regulation that will affect OEMs, suppliers and disruptors.

Room B Morning Day 2
09:00 - 12:55

Moderator

Dr Semih Bilgen
Dean of engineering and natural sciences
Istanbul Okan University
Turkey

09:00

MLOPS platform for synthetic and real data

Danny Atsmon
Founder and CEO
Cognata
Israel
ADAS and autonomous driving validation takes a very long time and generates a humongous amount of data (either synthetic or real). This data is usually left untouched not to mention any reuse, analysis or transformation. Cognata has developed a cloud-based MLOPS platform that gathers real and synthetic data, analyzes it and makes it reusable either as synthetic scenarios sensors using real-world data collection.

09:25

Breaking the fundamental robust perception gaps for ADAS and AV

Matthias Schulze
VP Europe and Asia
Algolux
Germany
ADAS and AV platforms require accurate computer vision (CV) to deliver safe operation. Unfortunately, as seen in recent reports from the US AAA, Tesla at CES 2022, and direct experiences from OEMs and Tier 1s, the significant gaps in the practical operating effectiveness in all conditions of the latest ADAS and AV systems deployed are disheartening. Algolux will present multiple proven best-in-industry approaches that are applied to optimize existing vision systems to significantly improve vision performance, rearchitect current methods to massively improve CV robustness in harsh conditions, and deliver unprecedented robust accuracy of dense depth perception.

09:50

Rethinking testing and validation in an environment of increased connectivity

Alexander Bodensohn
Director Business Development
Aurora Labs
Germany
New testing, reporting and validation methodologies are needed to answer both the challenge of the growing amount of regulated functions and the frequency of new software. Testing methodologies and technologies are required to embrace continuous testing throughout the vehicle’s lifetime, even on the road. Software quality maintenance is required to change from being reactive when an error has been detected to being proactive in detecting changes in the software functionality behavior before the error has occurred. A standard language using mathematical and technical methods is required to provide evidence of the changes in software for regulation.

10:15 - 10:45

Break

10:45

AI algorithms for ADAS – rule-based vs 4D AI

Kris De Meester
VP - sales and business development
XenomatiX
Belgium
In the development of ADAS and AD algorithms, AI is often understood to imply NN training and validation. But artificial intelligence exceeds by far the domain of NN programming and, depending on the use case, NN is not always the best choice. This paper presents state-of-the-art AI programming and distinguishes cases favoring NN from cases favoring a rule-based approach. Application examples like Highway Pilot, City Pilot and Traffic Monitoring are discussed, isolating differentiating features requiring different approaches. The paper focuses on pure point clouds from lidar as this is the more recent data domain with limited accomplishments in automotive apps.

11:10

Verification strategy of a complete Level 3/Level 4 AD stack

Mahmoud Serour
Autonomous driving team lead
Brightskies
Egypt
This talk presents ODD-centric verification of the perception, mapping, localization, path planning and vehicle control modules of a Level 3/Level 4 autonomous driving platform. The presented approach follows two levels of verification, component-based methods and a holistic approach for an exhaustive failure mode analysis. The diversity of the presented methods ranges between simulation-based models and deployment-based tests to ensure complete coverage of corner cases and realistic analysis of the algorithms. The findings presented are the results of challenges encountered during the deployment of a Level 3 robotaxi project at KAUST in Saudi Arabia.

11:35

Challenges of AI Software Testing: Understanding and Finding Its Limitations

Khanlian Chung
Senior software development engineer
Vector Informatik
Germany
Alexander Faul
Senior software development engineer
Vector Informatik GmbH
Germany
One of the most important issues regarding the integration of AI-based software in security-critical environments is how to test them. One cornerstone of AI testing could be explainable AI (xAI). It helps software developers understand the decision-making process of AI algorithms, which is fundamental for trustworthy AI. Another important aspect is to find the limitations of the security-critical software. Utilizing AI methods can complement conventional approaches and yield a more comprehensive risk assessment. In this talk, we show how we implemented xAI and AI testing methods in specific use cases.

12:00

Risk-aware autonomous vehicle perception

Chess Stetson
CEO and founder
dRISK
UK
How do we make a safe, fully driverless car? While AVs tend to perform adequately during normal driving conditions, they still fail on edge cases – the many hard, high-risk scenarios that are individually unlikely but collectively make up all the risk. In this presentation, we introduce a novel solution to training on edge cases, all the way to the level of the perception system, resulting in a 6x improvement in detection of high-risk events with 2x improvement in detection confidence. This makes AVs safer and thus commercially viable.

12:25 - 13:55

Lunch

Room B Afternoon Day 2
13:55 - 16:55

Moderator

Abhash Das
Senior safety staff
UL International Germany GmbH
Germany

13:55

An open innovation platform for autonomous vehicles software

Dr Semih Bilgen
Dean of engineering and natural sciences
Istanbul Okan University
Turkey
The Open Innovation Autonomous Vehicle Development and Testing Platform (OPINA) (www.opinaproject.com) is an EU-funded project that aims to establish a research, development and commercialization ecosystem for autonomous vehicles. Developers of all sizes, including researchers, startups, SMEs and OEMs will be able to integrate and test their software with other components to be provided on the platform, with strict protection of information privacy. Simulation and testing facilities, including MIL, SIL and HIL as well as driver-in-the-loop (DIL) systems will be available, together with on-vehicle deployment facilities for developed software, incorporating state-of-the-art sensor facilities.

14:20

The Autonomous Safety and Architecture Working Group

Christoph Schulze
Project manager
TTTech Auto AG
Austria
The Autonomous is a global community bringing together leading organizations to collaborate and advance safe autonomous mobility. Under its innovation stream, it launches and facilitates working groups made up of experts from diverse companies who work together to facilitate the adoption of autonomous vehicles on the road. The Safety and Architecture Working Group aims to help accelerate the market readiness and series development of safe self-driving cars by formulating a conceptual reference architecture for a Level 4 AD feature. We want to compare our choice against other candidate architectures and build a safety argumentation for the reference solution.

14:45

Affordable AI for automotive embedded applications – workflow and use

Florian Hauer
Embedded AI expert
Infineon Technologies
Germany
Autonomous vehicles leverage artificial intelligence in all their driving as well as connectivity functionality. This requires the seamless and cost-efficient yet reliable integration of embedded AI technology in all such applications. As a solution, we propose a combination of suitable hardware combined with an overarching software ecosystem that enables the deployment of neural networks for real-time applications. We present such an embedded AI workflow along the line of specific use cases, i.e. how it enables cost-efficient, real-time AI-based virtual sensors, e.g. for e-motor control, as well as AI-based prediction of physical quantities, e.g. for trajectory planning and control for vehicle motion.

15:10 - 15:40

Break

15:40

From AD prototyping to ISO 26262 production-grade software

Nicolas du Lac
CEO, CTO
Intempora
France
In just a few years, software has become a major keystone strategy for the automotive industry. The industry has realized that automated driving development is way more complex than initially expected and still represents a serious challenge in terms of performance and safety. This talk will explain how a data-driven development approach and a well-designed software framework can help speed up the development steps of autonomous vehicles with unprecedented performance, from the early stages of prototyping to testing, validating and deploying production-grade software, allowing developers to fill the gap from the code to the road.

16:05

Bridging ADAS to AD in mass production

Dr Duong-Van Nguyen
ADAS department manager
Panasonic Automotive Europe
Germany
The presentation will discuss the gap between current ADAS systems and expected ADAS and AD systems; technologies to enable 3D sensing using low-cost sensors by available ADAS ECU; advanced sensor fusion to compensate the deficiencies of one system with the abilities of others. It will also introduce a world-first L4 parking system that uses low-cost sensors.

16:30

Next-generation sensor fusion

Eric Richter
Director of technology innovation
Baselabs
Germany
Next-generation driving functions, like automated driving in urban environments or automated parking, are targeting an increasing number of highly complex scenarios with many different traffic participants and object types. Next-gen sensors, like high-resolution radars or cameras with semantic segmentation information, provide detail-rich information about the environment. Current sensor fusion approaches have shortcomings in dealing with the requirements that arise from such settings. We will present the next-gen sensor fusion technology, which provides information for static and dynamic objects as well as free space, with high quality and robustness, through an integrated sensor fusion approach.

Join your chosen session in the Autonomous Vehicle Test & Development Track

Join your chosen session in the Autonomous Vehicle Test & Development Track



Room A Validation, verification, data management and in-the-loop testing
09:00 - 16:55

Moderator

Holger Banzhaf
Managing director
DeepScenario GmbH
Germany

09:00

Comparability and exchangeability of scenario databases

Sytze Kalisvaart
Senior project manager StreetWise
TNO
Netherlands
Scenarios have become central to AV safety assessment to describe situations an AV will meet. TNO collects scenario statistics from driving data. These help to quantify a positive risk balance – is the new generation safer? Now that driving data collection is established at OEMs and tiers, the next step is comparability and exchangeability of such scenarios and statistics. TNO compares databases at the level of parameter distributions to identify biases or blind spots. This helps to direct data collection and completeness for legal due diligence. A consortium of industrial partners has started to exchange scenario statistics, coordinated by TNO.

09:25

AV beyond the hype: how to prepare for mass deployment

Robin Van der Made
Director product management
Siemens Industry Software
Netherlands
The concept of AV has been proven in many pilots during the last decade. The first companies are offering commercial AV services at a limited scale. Now, the industry needs to prepare for scaling up. These vehicles will serve new locations with new driving environments, such as more variable weather and road conditions, diverse driving cultures and further challenges. How do we deal efficiently with this increase, especially in the light of scarce engineering resources around the globe? This presentation will show how to grow the expansion of AV efficiently without growing the underlying development and validation resources at the same pace.

09:50

Generating synthetic data for neural network training

Daniel Tosoki
Product & operations manager of aiSim
aiMotive
Hungary
Real-world data is vital for training detection algorithms. However, gathering data on edge cases for training purposes is not always easy – sometimes it isn’t even possible. Furthermore, no matter how much training data is collected, it will always be unbalanced, and there are dangerous and complicated scenarios (for example, close car data) to be collected, which should and can be replaced by simulation. We will show on real examples and measurements how synthetic data generated by aiSim, the world's first ASIL-D certified validation tool, helped overcome these difficulties and improved the performance of neural network-based perception.

10:15 - 10:45

Break

10:45

HIL simulations reach a higher dimension of realism

Gordan Galic
Technical marketing director
Xylon
Croatia
Matt Daley
Operations director
rFpro
UK
Xylon and rFpro teamed up to explore methods of connecting hardware ECUs to simulation software. In this presentation, we will describe how we immerse a real-world surround-view parking assistance ECU in a simulated world seen by four fully modeled virtual HD video cameras placed on the vehicle model driving along virtual roads. This ADAS ECU was selected based on its highly visual nature and required transfers of large amounts of video data. We will outline the challenges and solutions for quick and precise translations between physical and virtual domains that enable this full-speed, closed-loop HIL simulation setup.

11:10

Do you have the right data for AV development?

André Rolfsmeier
Director - strategic product management
dSPACE GmbH
Germany
Petabytes of data are recorded during AV development and the costs for managing the data are immense. In this talk, we will explain how to explore, tag and manage huge amounts of data efficiently. By means of AI-based detectors, you will gain a deep understanding of what data you already have and what data is still needed to develop and test safety-critical perception and autonomous driving systems. Using our own data acquisition vehicle, we collect 30+ TBs of data per test drive. Our experience in automatically detecting defined traffic scenarios in real time during driving or in post-processing will be shared.

11:35

Virtual assessment framework and ontology-based test case generation methods

Dr Jianbo Tao
Senior lead engineer ADAS/AD XiL simulation
AVL List GmbH
Austria
Dr Heiko Scharke
Senior global product manager for ADAS/AD solutions
AVL List GmbH
Austria
A comprehensive and reliable testing process for automated and autonomous driving functions must follow a reasonable path from parameter selection and test scenario generation to test execution and final assessment of the achieved test coverage. In the presentation we will provide an overview of our ontology-based testing approach that is based on five main pillars: the scenario sources, the ontology model setup, the automated test scenario generation and planning, the test execution and evaluation framework and the test management system. The test management system ensures continuous monitoring of the testing.

12:00

1+1=3: hybrid HIL systems for cost-effective ADAS validation

Johann Führmann
Head of business development
b-plus automotive GmbH
Germany
Due to the UNECE-VMAD working party, hybrid HIL systems can now be used to mix real-world driving with simulated driving. At first glance, a lot of validation expenses can be saved by this, but many of the pitfalls will stay the same. Setting up HIL clusters can be time-consuming and expensive in addition to requiring a lot of space in the data centers. Reusing existing SIL hardware can decrease the initial expenses significantly. A thoughtful choice of farm location, test mixture and maintenance strategy will ensure cost control right from the start.

12:25 - 13:55

Lunch

Moderator

Dr Clara Marina Martinez
Engineer - ADAS virtual development
Porsche Engineering Services
Germany

13:55

We keep you safe on the road (dynamic risk management)

Dr Thomas Freudenmann
Co-CEO/co-founder
EDI GmbH - Engineering Data Intelligence
Germany
The RELAI – Risk Estimation with a Learning AI – project's approach is to analyze real-world driving scenarios and the driving behavior necessary for safe driving, which takes into account not only the official traffic rules but also the traffic behavior of other road users and generates test scenarios from this data. The made system in the RELAI project is backed up by the dynamic risk management (DRM) system, which enables autonomous driving systems to dynamically manage various risks in road traffic as experienced human drivers do. This allows the autonomous vehicle to make important decisions based on the driving context.

14:20

Real-time over-the-air automotive radar hardware in the loop to test autonomous vehicles

Dr Kasra Haghighi
CEO
Uniquesec
Sweden
Fahimeh Rafieinia
CTO
Uniquesec
Sweden
Autonomous driving will revolutionize the future of mobility. The main pillar of autonomous driving, as well as advanced driver assistance safety systems (ADAS), is sensing. Among all sensors, radars are the most reliable and versatile ones, providing environment perception. Test and validation of radars and radar-based safety functions is a necessity that requires an advanced OTA radar HIL setup where rich and dynamic traffic scenarios can be emulated. Certain real-time requirements need to be established between environment simulator and radar-target simulator. This presentation will demonstrate a full OTA HIL setup enabling emulation of more than 200 moving targets for the radar under test.

14:45

Left-shifting the validation of autonomous driving using simulation

Jihas Khan
Senior specialist
Tata Elxsi
India
Simulation is an ideal choice for the left-shifting of AD validation, ensuring quality, cost and time savings. Software-in-the-loop testing, synthetic scenario simulation, virtual ECU development and hardware-in-the-loop testing are various types of virtual validation of AD, which will be explained with use cases in the presentation. However, ensuring process compliance and uniform data exchange is of paramount importance for these virtual simulation scenarios. The presentation will also touch on how the ASAM open framework can play a vital role in ensuring the uniformity of test and simulation frameworks. The presenter will show case-relevant live work samples.

15:10 - 15:40

Break

15:40

How can we better streamline test and validation processes?

Simon Tiedemann
Senior product manager - EB Assist Test Lab
Elektrobit
Germany
Bastian Speth
Senior business manager
Elektrobit
Germany
We face many challenges in how we record and consume data. We’re finding inefficiencies in the process with incomplete or unknown data. Elektrobit has proven expertise in data collection and processing, with hardware and software used in millions of vehicles today. Are storage costs increasing with each project? While conducting test drives all over the world, Elektrobit has developed innovative solutions that have saved customers significant time and money. This presentation will share how a new data orchestration platform can help collaboration and data socialization, connecting data scientists, managers, engineers, testers and algorithm developers to achieve common goals.

16:05

Sensor fusion simulation using 3D imaging radar and lidar

Koki Taniyama
Engineer
OTSL Inc
Japan
Recently, 3D imaging radar performance has been increasing, and this point cloud data will catch up with the lidar point cloud in the near future. OTSL has developed the new MIMO Radar Simulator (COSMOSIM-AMMWR2) for supporting 3D imaging and point cloud, and has implemented various Tier-1 dedicated radar models for more accurate simulation. The presentation will showcase show genuine MIMO radar simulation in real time.

16:30

A distributed simulation for autonomous vehicle test and development

Amir Soltani
Founder and managing director
KAN Engineering Ltd
UK
The critical role of simulation in the successful development, test, verification and validation of connected and autonomous vehicles is now being widely acknowledged. Still, no single simulation tool is available that can cover the entire development V-cycle. Under Zenzic's interoperability projects, KAN Engineering has delivered a novel solution for an integrated end-to-end distributed interoperable simulation architecture across the UK. The developed technical solution enables seamless deployment between various stages of development from MIL, SIL, HIL, DIL, VIL, test tracks and real-world roads, to carry out similar tests in different simulation contexts and geographical locations, all from a single place.

Room B Connectivity, mapping and positioning
09:00 - 11:35

Moderator

Benjamin Engel
Global technology manager
ASAM eV
Germany

09:00

Mapping change: new approaches to map maintenance for AD applications

Nicola Consolati
Head of OEM delivery
Woven Planet
Japan
Map maintenance - the updating of maps to account for changes in the road - has long been one of the most challenging aspects of automotive mapping, especially for automated driving applications. This complexity is becoming ever greater as automated driving technology proliferates and accurate, reliable automotive-grade mapping is needed at continental scale. In this presentation, Nicola Consolati will help size this ‘change’ problem and discuss new, multimodal approaches to managing change, using examples from recent work. The presentation will pay particular attention to the challenges presented by urban mapping and the technical solutions needed to map cities at scale.

09:25

V2X technology developments, testing requirements and certifications

Thomas Jaeger
Senior vice president
Dekra SE
Germany
The presentation will cover the latest V2X technology developments (ITS-G5, DSRC-wave, cellular), the latest regulatory and certification situation (EU, US and others), testing requirements for connected and automated driving (regulatory, conformance, use cases, interoperability), the most recent developments in regional and global interest groups featuring certifications, and challenges and the outlook for the future.

09:50

Distributed Mobility Cloud: From safety to automated driving support

Uwe Puetzschler
Head of Car2X
Nokia Mobile Networks
Germany
Starting with sharing of safety related traffic information the automotive ecosystem will have to evolve and move to more advanced use cases supporting automated driving and VRU protection. Edge computing will play an important role for the deployment of those advanced use cases. Experience gained during test deployments and the role of data for sustainable cooperation and business models will be presented and discussed.

10:15 - 10:45

Break

10:45

Assessment of GNSS-based equipment in the context of AEBS-UNECE R152

Ernst Phillip Mrohs
Head of laboratory
NavCert GmbH
Germany
With regulations in the areas of the pan-European-eCall, EETS and digital tachograph, obligatory requirements for GNSS-based in-vehicle systems are mandated. Similar is specified in the field of driving assistance, e.g. UNECE-R152. Conformity is verified in accredited laboratories. In each laboratory, the calibration of measuring instruments is essential for testing reliability. Since GNSS observations of such instruments are not directly traceable to SI units, an assessment, validation and qualification must be performed. For this, NavCert developed a procedure based on existing standards. The core is performance verification. An exemplary assessment of the PVT solution is presented here.

11:10

Vehicle platooning with dynamic performance optimisation

Tim Edwards
Chief engineer - connected autonomous vehicles
Horiba MIRA
UK
Vehicle platooning presents an opportunity to leverage automation and connectivity in combination to operate safely and efficiently in highway driving. The PLATONIC project, a collaboration between HORIBA MIRA and GMV NSL, funded under the ESA Navigation Innovation and Support Program (NAVISP) stimulating innovation in the PNT domain, addresses advances in cooperative positioning and integrity algorithms for platooning applications. This presentation will describe the project approach with a focus on the design and validation challenges, and how real-time integrity monitoring can be used for dynamic optimisation of vehicle speeds and following distances.

Room B Best practices, standards and regulations for the deployment of safe ADAS and autonomous driving technologies
11:35 - 16:55

Moderator

Benjamin Engel
Global technology manager
ASAM eV
Germany

11:35

The latest update on connected and autonomous driving regulations: global overview and implications

Alexander Schwab
Officially recognized expert for motor vehicle traffic
TÜV SÜD
Germany
With UNECE R157, an international framework was created in 2020 for the approval of an automated lane-keeping system (according to Level 3). In 2021, Germany introduced the possibility of permitting Level 4 vehicles on public roads by adapting the Road Traffic Act (StVG). The corresponding draft directive is in the final adoption process. This session will give an overview of the current regulations worldwide. We will present the basic ingredients of effective approval, considering virtual and physical tests, SOTIF and cybersecurity, using the example of the German Level 4 Regulation, highlighting the challenges and possible approaches to solving them.

12:00

Reference measurement system for development and validation of ADAS functionalities

Steffen Metzner
Product Manager ADAS Reference Measurement Systems
AVL List GmbH
Austria
Despite all efforts to carry out tests of vehicle automation functions using simulation technology, the major part of tests and the validation is still necessary to be executed directly in-vehicle. For an efficient development and verification of sensor and perception systems a highly accurate ground truth reference sensor device is necessary. Without such systems, troubleshooting and objective evaluation of vehicle behaviour becomes very expensive and often even not possible. The solution discussed in this presentation aims at supporting the development of vehicle automation functionalities during the entire development process from sensor selection to function validation.

12:25 - 13:55

Lunch

Moderator

Adrian Bertl
Teamleader product marketing
b-plus technologies GmbH
Germany

13:55

A new measure of AV Safety: Are existing measures good enough?

Dr Erik Antonsson
Founder and CTO
Streetscope Inc
USA
A new quantitative measure of safe operations in traffic, based on continuous measurement of the positions and movements (kinematics) of traffic objects has been developed and is being deployed. This new approach provides a dense, continuous, inexpensive, assumption-free way to measure, assess, and compare driving behavior, e.g., different builds of an AV developer’s software stack, or different configurations of sensors, and overcomes important problems with existing measures such as TTC, ISM, RSS, etc. A quantitative measure of this type is essential to measuring engineering progress and to providing an independent, quantitative basis for safety standards and public acceptance of AVs.

14:20

Tactical safety and its implications for autonomous vehicle testing

Hans-Peter Schöner
CEO
Insight from Outside Consulting
Germany
Situation awareness allows good human drivers to judge when to be particularly cautious, on top of following standard driving rules, deliberately undertaking additional actions such as reducing driving speed and increasing distances to other traffic participants. This behavior, denoted as tactical safety, allows the avoidance of constellations that could not be controlled safely, or that could hinder fluid traffic movement. When applying tactical safety strategies to autonomous vehicles, communication between the vehicle and other traffic participants and/or with the infrastructure is of particular importance. Finally, to enable the technical implementation of tactical safety strategies, the development of quantitative measures for the danger of specific situations is a key issue. This presentation will elaborate on tactical safety and its implications for autonomous vehicle testing. Co-Author Dr. Jacobo Antona-Makoshi, Japan Automobile Research Institute (JARI), SAKURA project

14:45

Level 2-3-4-5 overview of regulation/Euro NCAP status and testing challenges

Alain Piperno
Senior expert - autonomous vehicles testing and homologation
UTAC
France
UTAC is a worldwide reference in testing and type approval, an accredited lab for Euro NCAP tests and for French type approval tests. This presentation will review today’s regulation and testing capacity proposals and new challenges. It will cover Level 2-3-4-5 regulations status for M1 vehicles, shuttles, buses, delivery robots, SW/OTA, cybersecurity, AI and virtual type approval. We will also discuss needs/solutions for testing and validation, including physical, new protocols, new targets, rain and fog testing, digital and VIL testing, AI, V2X and cybersecurity.

15:10 - 15:40

Break

15:40

The law on autonomous driving in Germany – last-mile issue

Jörg Kahler
Attorney / partner
GSK Stockmann
Germany
With the newly enacted Autonomous Driving Act, Germany created the world’s first legal framework for the approval of SAE Level 4 vehicles out on public roads in specified operating areas. The legal certainty now provided is intended to bolster trust in and acceptance of these new technologies. The presentation will introduce the new law and its particular importance for the last-mile issue. It will also outline the remaining challenges regarding the issues of responsibility, liability and data privacy, and how to deal with them.

16:05

Challenges of testing the connected automated/autonomous vehicle – a holistic approach

Dr Alain Vouffo
Senior product manager automotive
Spirent Communications plc
Germany
The connected automated vehicle (CAV) is a system of systems combining a wide variety of wired and wireless communication technologies with different types of sensors to provide advanced applications for enhancing driving comfort and safety for vehicle owners and other road traffic participants. This presentation will discuss the key challenges facing the automotive industry in testing the CAV, along with some potential methods and tools for a holistic approach to addressing those challenges.

16:30

AD management system to fulfill technology and legal compliance

Lucas Bublitz
Senior consultant autonomous mobility / product and process compliance and PhD student for autonomous driving
P3 Automotive GmbH | University of Aveiro
Germany
This presentation will introduce a holistic AD management system (ADMS) framework that defines the safety, security and quality frame in terms of processes, methods, tools and organization through the whole digital lifecycle of AD systems. The framework meets the requirements of the safety and security norms like SOTIF, functional safety, cybersecurity and Automotive SPICE and ensures advanced risk and threat management, which will be continuously adapted, and change/configuration management, which requires fast incident reaction and function validation. An ADMS addresses the challenges of operations of SAE L3 and L4 applications considering the whole development phase, and sets the organizational structure.