Preliminary Conference Program

Autonomous Vehicle Test & Development Conference

Achieving safe autonomous driving through test and development, collaboration, harmonization and standards

HEADSTART project: Harmonized European Solutions for the Testing of Automated Road Transport

Álvaro Arrúe
Project manager connected and automated driving
HEADSTART is an H2020 EU-funded project that aims to define testing and validation procedures of connected and automated driving functions including key technologies such as communications, cybersecurity and positioning. The tests will be in the simulation and real-world fields to validate safety and security performance according to the key users’ needs. The expected impact of the HEADSTART project is based on three main action pillars: testing and validation – potentiation of development strategies bringing time and costs reduction; assessment – creation of assessment protocols increasing vehicle safety awareness; certification – support of regulations ensuring the safe introduction of CAD technologies to the market.

Safety assessment of ADAS using simulation with reliability analysis methods

Maximilian Rasch
PhD and engineer
Mercedes-Benz AG

New approaches for autonomous vehicles certification: a European perspective

Maria-Cristina Galassi
Scientific project officer
European Commission
The presentation will address AV safety verification through audit and assessment of the documentation provided by manufacturers (including simulations), physical testing (both on-track and on-road) and operational feedback from real-life experience.

Validating ADS toward an industrial scenarios database

Marc Pajon
Expert leader - testing and measurement technologies
Renault Group
Validating AV safety is a crucial part of ongoing research. The approach of separately track-testing sensors and driving algorithms is hardly sufficient to demonstrate AV safety. Scenario-based simulation approaches are necessary complements to the traditional approach, allowing computation of a controlled diversity of key variables in many iterations in a safe, fast and documented way. French car manufacturers Renault and PSA, together with academic researchers (VEDECOM, SystemX, Lab and Ceesar) and other partners (Valeo, AVS and Expleo), propose to address the challenge of demonstrating AV safety by taking an array of 'in-the-field' situations into account.

ASAM simulation standards – past, present and future

Benjamin Engel
Global technology manager
In 2018 ASAM acquired its first simulation standards in the form of the OpenX portfolio (openDRIVE, openCRG and openSCENARIO), with the addition of the Open Simulation Interface in 2019. Since then, simulation experts worldwide have been working hard on the first official ASAM revisions, the first of which are to be released in Q1 of 2020. This presentation will give an overview of the ASAM activities to date, and provide some insight into where we see the road leading in 2020 and beyond. Our goal is standardization to facilitate the development of safe, regulated autonomous driving.

SOTIF in a Volvo cars context

Carina Björnsson
Technical expert, driver assistance and active safety test methods
Volvo Car Corporation

Homologation of automated driving functions: worldwide overview, customer acceptance and strategic aspects

Christian Gnandt
Vice president automated driving
TÜV Süd Auto Service GmbH
Homologation of automated driving functions presents a huge challenge for their market introduction. Existing regulatory safety frameworks applicable to conventional vehicles and their components are insufficient to fully assess the operational characteristics of current and future automated vehicle technologies. With increasing automation, vehicles transform into cyber-physical systems that no longer require a human driver; therefore, new safety challenges will have to be considered. This presentation discusses those challenges, provides an overview of the current regulatory and standardization work in progress and explains the possibilities for how to approve automated vehicles for public roads today.

From absolute safety to informed safety: the role of operational design domain

Dr Siddartha Khastgir
Head of verification and validation, intelligent vehicles
WMG, University of Warwick, UK
To prove that automated driving systems (ADS) are safer than human drivers, it is suggested that they need to be driven for over 11 billion miles. A number of miles is not an appropriate metric and doesn’t guarantee absolute safety. This highlights the question: “How safe is safe enough?” To answer, we suggest a departure from the world of absolute safety to informed safety. A key aspect of informed safety includes an accurate and standardized definition of the operational design domain for an ADS, including conveying it to users, regulators and other stakeholders. The first step in all safety standards: ODD definition.

Coordination of R&I and pilot activities in Europe: how can we learn in order to scale up?

Stephane Dreher
Senior manager - connectivity and automation, blockchain
Ertico - ITS Europe
Many research and innovation, testing and piloting activities are being carried out independently across Europe either through EU-funded consortia projects or at the national level driven by Member States, industry or public-private partnerships. The lack of alignment and common approaches often results in duplications or overlaps, prevents comparability of results and the development of interoperable solutions, and hampers the harmonized deployment of CAD in Europe. In an effort to facilitate the exchange of lessons learned and best practices, as well as the identification of synergies and gaps between projects, the EU-funded ARCADE Coordination and Support Action has set up a comprehensive knowledge base on CAD-related activities in Europe and beyond, which is currently being populated in collaboration with the EU CCAM Single Platform. The project is also organizing concertation events to support consensus-building across stakeholders on methodologies or the identification of next R&I actions on CAD. This presentation will provide an overview of the CAD knowledge base and key outcomes from other harmonization-related activities of the project on methodologies or description of use cases.

Humanizing autonomy: setting a global standard for how autonomous systems interact with people

Raunaq Bose
Co-founder and CTO
Humanising Autonomy
By building a human intent prediction software, we are able to increase VRU safety while also improving public acceptance of AVs.

Better safe than sorry – safety assessment for self-driving vehicles

Mario Torres
This presentation outlines a method for defining safety measures and provides concrete examples of such measures. Safety measures are a vital element for validation of the behavior of highly automated vehicles – that is, for the assessment of the behavior of such vehicles from an observer's point of view. Validation complements the verification process, which studies the correct functioning of a system and its components compared with their requirements.

Assurance cases for automated driving

Rasmus Adler
Program manager
Fraunhofer IESE
This presentation covers challenges, solutions and standards for arguing the safety of automated driving. In the V&V Methoden project, Fraunhofer IESE is working on a safety argument that can be based on evidence from testing and simulation. The talk will summarize the current results of this project. It will also discuss UL 4600 and the upcoming application guide of the AK DKE/AK 801.0.8 Spezifikation und Entwurf autonomer/kognitiver Systeme (specification and design of autonomous/cognitive systems).

Validation and verification methodology for AI in autonomous vehicles

Leon Altarac
VP business development
The appearance of autonomous vehicles (AV) on public roads is promised to be a reality soon, but some critical aspects of this reality are yet to be resolved. One of these is the lack of an efficient safety performance verification technique, as the existing tools for hardware and software reliability and safety engineering do not provide a comprehensive solution regarding algorithms that are based on machine learning (AI). To start tackling this problem, a methodology based on statistical testing in a simulated environment is presented and demonstrated on full-scale autonomous vehicles.

Mapping, positioning and connectivity

Bridging the gap between AV navigation software and hardware

Greg Drew
Polysync Technologies
Currently there is a gap between autonomous vehicle (AV) navigation software and physical operation. This presentation will discuss the current status, challenges, pitfalls and best practices for integrating AI and physical navigation systems and ensuring autonomous vehicle safety. The business results of achieving the ability to write once, run anywhere include efficiency, safety, fastest time-to-market, normalization and best-practice optimization. A universal safety standard for integrating software and hardware navigation systems will streamline and normalize the development and optimization of AV navigation systems, accelerate market entry and achieve the promise of AVs to dramatically improve safety.

The collaborative way forward: open data for an autonomous future

Emil Dautovic
Vice president automotive
For autonomous vehicles to safely navigate our roads, they need meticulously detailed and accurate information about the world around them. No single actor is capable of collecting all the data necessary for these vehicles to operate in all locations and situations, so the only way forward is to open and share data. This collaborative approach has driven the growth of roughly one billion street-level images on the Mapillary platform – all used to create data sets for training machines to see and understand the world, and to build and maintain the high-definition maps required for autonomous vehicles to be deployed globally.

The challenges in testing connected vehicles

Stoyan Nikolov
Test analyst
McLaren Applied Technologies
The complexity of connected and autonomous vehicles increases significantly with the introduction of multiple connectivity channels and sensors. The requirements for robust and uninterruptible connection provoke the need for multiple modems fixed to multiple network operators, being able to switch among various wireless networks (such as 3G, 4G and 5G). The introduction of geofence-triggered software features requires GPS/GNSS connectivity in addition to the mobile network. Testing the connectivity is a challenging task that requires realistic simulation of the mobile network conditions and GPS/GNSS. This presentation will cover the challenges of testing the connectivity channels for connected vehicles, the simulation scenarios to be considered and the challenges of simulating a fleet.

On-track testing of connected vehicles: methodology, challenges and results

Annie Saleh
Head of automated and connected driving
PMG Technologies
PMG Technologies has completed closed-circuit track testing for Transport Canada to evaluate the performance and effectiveness of crash avoidance technologies, specifically connected vehicle technology. Tests were performed using two DSRC-equipped vehicles to reproduce maneuvers that would trigger targeted V2V features (hard braking, slippery road and disabled vehicle). The test results are used to describe and quantify the timing of visual and audible alerts sent to the driver after the reception of basic safety messages. This presentation discusses the testing methodology and challenges of physical testing on tracks. It highlights the analysis of test results and demonstrates the importance of track testing.

Pedestrian protection through automated driving

Hartmut Runge
Project manager DriveMark
A new concept for an autonomous driving system is presented, which in particular protects other road users such as pedestrians and cyclists. Furthermore, the system automatically ensures compliance with traffic regulations in a smart city equipped with it. It goes far beyond what we have seen so far with lane-keeping or automatic braking systems. The concept of geofencing will be further developed with unprecedented granularity. Detailed maps of the traffic areas of a city with 4in resolution will be used.

Automated public road testing based on digital twins

Patrick Luley
R&D manager - automated driving
Joanneum Research - Digital
To pave the ground for salable and cost-efficient test and validation of AD functions by real testing on public roads, Joanneum Research is producing Ultra High Definition Maps (UHDmaps) based on mobile mapping data in a salable automated workflow. UHDmaps contain a digital copy of reality, which sets the benchmark for the digital assessment of automated driving functions. The depicted solution is already utilized by the Austrian Light Vehicle Proving Region for Automated Driving (ALP.Lab GmbH) and its partners. The presentation will give an overview of the technical solution and certain test use cases.

Precise HD map data as the basis for virtual testing and simulation

Dr Gunnar Gräfe
3D Mapping Solutions GmbH
Artificially designed digital roads may help case by case, but for various applications the precise digitization of real-world roads is needed. The typical requirement is that the roads used for virtual testing and simulation are regarded as identical digital twins, which is a prerequisite for comparable testing in reality and virtual environments. 3DMS has invented the necessary technology for more than 20 years and generates high-resolution digital road-surface models in OpenCRG format, or produces precise high-definition reference maps in OpenDRIVE format, which are either used for virtual simulation and testing or as reference maps in the car for autonomous driving development. The presentation will show various project examples.

Managing the test and development process – best practices for accelerating development and achieving safe autonomy

Building a toolchain and an ecosystem – the strategy behind creating a new company for autonomous solutions

Magnus Liljeqvist
Global technology manager - infrastructure
Volvo Autonomous Solutions

Scaling a simulation toolchain for higher levels of autonomy

Dr Sandeep Sovani
Global director, automotive industry
Ansys Inc
Driving automation software for ADAS Levels 1-2 is now routinely validated by software-in-the-loop (SIL) simulations. Moving up to Level 3 greatly increases the complexity of the automated driving system, thus increasing validation needs exponentially. This talk presents the ANSYS Autonomy toolchain for virtual validation and sign-off of automated driving features at Levels 3-4. For these levels, beyond SIL simulation the toolchain includes software solutions for scenario collection via drive data analytics, scenario curation, scenario variation, test plan management, robustness testing of perception software, simulation result analytics for coverage analysis and building statistical validation cases, simulation data management and toolchain validation.

Product management for autonomous vehicles

Todd Medema
Product manager
Uber ATG
Most product management resources are focused on B2C or even B2B. But what does it take to design and build internal tooling to accelerate autonomy development? In this talk, we'll cover some of the unique challenges in products built by engineers for engineers, capable of handling immense datasets and complexity, in an unsolved and ever-changing solution set.

Ensuring successful implementation and uptake of new mobility services

Kevin Vincent
Director - Centre for Connected and Autonomous Automotive Research
Coventry University
The technological challenges/opportunities of CAV/CAM and the roadmap to autonomy are maturing. However, the requirements to generate market pull and the skills required for the successful implementation and uptake of services are less well understood. For example, safe ongoing operation of vehicles needs new MOT tests addressing security, software and data privacy; new and disruptive ownership models present issues and opportunities regarding individual design and brand identity; trust and perception require more human-centered design for viable solutions. These and more questions need answering if economic, environmental and productivity projections are to be realized.

Autonomous Vehicle AI, Software and Sensor Fusion Conference


Bridging ADAS to AD in mass production

Dr Duong-Van Nguyen
ADAS department manager
Panasonic Automotive Europe
The presentation will discuss the gap between current ADAS and expected ADAS and AD systems. It will also outline technologies to enable 3D sensing using low-cost sensors by available ADAS ECU, and examine advanced sensor fusion to compensate for the deficiencies of one by the abilities of others.


Automotive Grade Linux: enabling industry collaboration through open-source software

Dan Cauchy
Executive director, Automotive Grade Linux
The Linux Foundation
The Automotive Grade Linux (AGL) community consists of more than 150 companies across the automotive and tech industries who are working together to develop an open-source software platform 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 talk provides an overview of AGL, production use cases including Toyota and Subaru, the project roadmap and how to get involved.


ADAS/AD virtual end-to-end software development

Dr Clara Marina Martínez
Engineer - ADAS virtual development
Porsche Engineering Services GmbH
ADAS/AD software development needs to cope with complex sensor systems, plentiful corner cases still to be discovered and a cumbersome number of kilometers to test/certify. These tasks require high support from virtualization to be achievable under challenging deadlines and at reasonable cost. The perfect tool that gathers all your requirements does not exist. However, many high-quality software solutions are able to simulate sensors, traffic, vehicle dynamics, driver behavior and realistic environments with the level of detail that every project needs. At Porsche Engineering, we bring together the best tools, data sources and our experience in automotive, to create a flexible simulation platform to support end-to-end ADAS development.


Autonomous driving and open-source technology – does it fit?

Andreas Riexinger
Product manager
Robert Bosch GmbH
Automated driving solutions introduce a new complexity into the development of embedded systems in cars. This complexity rises with each level of control and autonomy. The toolchain for such challenges is also complex and the integration of all the tools requires considerable effort without a real competitive advantage for the automated driving solution. Instead of solving these challenges alone, wasting lots of money along the way, Bosch's automated driving division has started an open-source community known as OpenADx. This talk will present the open-source approach, the current state of the community and the currently available solutions.


Building an ADAS test and development environment in the cloud

Gabriel Sallah
EMEA HPC and big data architect, autonomous driving platform solutions
United Arab Emirates
This session will focus on the key Azure Cloud services needed to meet the demanding end-to-end requirements of testing and validating autonomous driving vehicles: from large-scale data ingestion (PB), to large-scale simulations (60,000+ cores) using high-performance computing (HPC), to scalable machine learning model creation, deployment and management. The presentation will share real-world experience of successfully building this platform for major OEMs and tier suppliers.


Use of artificial intelligence in the validation domain

Ahmed Yousif
Software design engineer
Valeo Schalter und Sensoren GmbH
The presentation will talk about the challenge of annotation and how it is solved using machine learning and AI. It will include a demo related to the topic and also some videos and deployment examples.


The future of driving behavior in autonomous vehicles

Davor Andric
CTO AI and analytics North and Central Europe
DXC Technology
The vision for fully autonomous vehicles has yet to be realized. How realistic is it? Despite the increase in commercially available autonomous features up to SAE Level 3, achieving Level 5 autonomy will require a very different development approach. We will review current approaches and challenges for autonomous driving development, including human driver behavior, and examine what is needed to develop autonomous driving technologies for intelligent and safe real-time driving.


Paving the way for autonomous driving

Bryan Berezdivin
Autonomous systems
Connected, autonomous, shared and electric vehicle trends are converging to revolutionize the automotive industry. In this unprecedented age of innovation, automotive companies rely on Amazon Web Services (AWS) to fuel their digital transformation efforts and get their products to market more quickly, while retaining ownership and control of their data and brand experience. Learn what challenges autonomous driving is posing for developers, and how the main players in the industry are addressing those challenges.


The AV test fleet of the future is virtual

Serkan Arslan
Director of automotive
Nvidia EMEA
Self-driving cars can improve safety while giving people the freedom of mobility. However, validating a self-driving vehicle solely using physical on-road testing is an insurmountable task. As humans, it’s easy to identify situations in which we’re accident prone, but it’s impossible to train for rare, dangerous scenarios/extreme conditions, or at large-enough scale. Through advances in AI and accelerated computing, simulation has emerged as the ideal solution for safely testing and validating AVs. Learn about a unique virtual AV test fleet in the cloud that will enable AV development and validation without putting others on the road in harm’s way.


AI and big data management for autonomous driving (AD)

Frank Kraemer
Systems architect
Advanced driver assistance systems (ADAS) are already becoming part of all vehicles today, and fully autonomous driving (AD) is in the development phase right now. To win this race, applied artificial intelligence (AI) is the key component. All major OEM and Tier 1 auto manufacturers are implementing and testing AD facilities. Developing and testing AD systems requires the storage and analysis of more data now than ever before. Researchers and developers who can deliver insights faster while managing rapid infrastructure growth will be poised to be industry leaders.


ADAS development and validation workflow and methodology

Dr Florian Baumann
CTO automotive and AI
Dell Technologies
IT and AI are key components of your development toward autonomous driving and the next generation of ADAS. Using an efficient workflow, you can make your engineers extremely productive and happy. This session will introduce you to the complete workflow of AI-enabled ADAS/AD product development.

Please note: this conference programme may be subject to change


Autonomous Vehicle International magazine