Preliminary Conference Programme

Day 2: Wednesday 22 May

Room A V2X, 5G & Connectivity
09:00 - 13:40


Bringing 5G to market requires innovative products and strong partnerships

Mike Peters
Executive vice president and president connected car division
Harman International
In the era of the Internet of Things (IoT), connectivity is a foundational element of next-generation connected and autonomous vehicle technology. It is an enabler for improved vehicle safety and an enhanced user experience. Consumers, regulators and law makers need to trust that these technologies are reliable, secure and safe. To achieve this, the upcoming 5G network infrastructure will provide best-in-class connectivity that will deliver a range of enhancements including higher bandwidth, lower latency and improved coverage. 5G is quickly emerging as the preferred industry solution for the next decade. Bringing 5G to market requires experience, innovative products and strong partnerships.


Autonomous driving: how to network complex IoT landscapes?

Oliver Bahns
Head of business area connected mobility
T-Systems International GmbH
Millions of connected vehicles and traffic infrastructure elements generate huge amounts of data that must be managed and interpreted as one IoT network. Only by an end-to-end operation of this complex and safety-relevant overall system will autonomous driving be feasible in practice. In many current research and development projects, individual 'solution islands' are being created. To realize practical use cases, these isolated experience domains and working models must be networked to create IoT landscapes with a multi-partner, multi-vendor and multi-technology view. This should include legal, compliance and security aspects in addition to technical categories such as stability, availability and performance.


Intelligent intersections – solving connected vehicle safety

Dr Ryan Monroe
Street Simplified
There is a limit to what a vehicle can see. The majority of autonomous vehicle collisions occur at intersections. How can intelligent infrastructure help? The presentation will explore case studies of AV collisions and why they happened, and look at the broader trend of collisions in general and what sets AVs apart. It will examine real videos of near-miss events, and give examples of common accident types, why they happen and how they can be predicted (using computer vision). Attendees will see how to use CV on the infrastructure to map where all road users are, forecast potential safety issues and communicate to enhance safety for all road users.


Understanding V2X application testing with 3D ray tracing multi-path GNSS simulation

Ali Soliman
Technical product manager, connected autonomous vehicles testing
Spirent Communications
This presentation relates to V2X application performance testing with 3D ray tracing multi-path GNSS simulation. V2X systems use accurate GNSS positioning to provide precise position information for safety. The speaker will discuss how multi-path can create errors in a GNSS receiver within urban environments, and explain how errors can vary due to the geometry of satellites and environmental conditions. The presentation will also explain how testing is crucial for the success of V2X technology, and will include details of a novel system that combines V2X emulation with an advanced GNSS simulator and propagation model, which depends on a 3D reconstruction of real life.


A test concept for testing V2X vehicles on test fields

Dr Fatih Ozel
Project manager
Oecon Products and Services GmbH
Cooperative intelligent transportation systems using vehicle-to-everything (V2X) communications provide opportunities for achieving sustainable transportation. However, before any V2X hardware and/or software infrastructure is fully developed and deployed, an extensive phase of testing is required. In this presentation, a test concept, which will be implemented in a motorway section of the A39 near Brunswick, is described for testing V2X systems on test fields by emulating cooperative behavior for normal vehicles in real traffic and in the vicinity of a cooperative test vehicle. This is achieved by utilizing systems such as stereo cameras and roadside stations.

11:05 - 11:35



HiL testing of 802.11p and C-V2X for the connected car

Axel Meinen
Technical sales manager
S.E.A. Datentechnik GmbH
A key issue for highly automated and autonomous driving is the connectivity between the car itself and the surrounding infrastructure via 802.11p and cellular V2X (LTE-V). Due to the safety-relevant functions of most of the components and applications, the mandatory and complex test and validation processes require an open and flexible platform that allows developers to access, analyze and manipulate signals and data in the various implementation phases within the development process. This presentation gives a detailed overview of the various testing approaches and highlights the advantages of an open testing platform.


Vehicle-to-everything technology (V2X) – the key enabler for autonomous vehicles

Alain Vouffo
Product manager - automotive
Spirent Communications
Vehicle-to-everything technology (V2X) is widely accepted as a key enabler for the autonomous vehicle. Beside the IEEE802.11p (a.k.a DSRC)-based V2X technology, cellular-V2X (C-V2X) has been gaining momentum. C-V2X provides vehicle-to-everything communication over the cellular network, based on 3GPP standards like LTE-V and 5G. Despite similarities with 11p-V2X, C-V2X comes with a range of specific test requirements and challenges that must be addressed early for this technology to deliver the expected benefits. This presentation discusses test requirements and challenges in addressing those. It highlights testing methodologies for C-V2X that ensure that new products and solutions are delivered in time.


Cooperative driving function development and testing in a virtual environment

Dr Natalya An
Business development manager ADAS and automated driving
IPG Automotive GmbH
Cooperative ADAS and AD functions, also called V2X applications, rely on various sensor data and reliable wireless communication. Virtual environment enables reproducible, efficient, risk-free testing of V2X applications. During early development stages, neither real radios nor virtual vehicle prototypes are necessary. To study the impact of network protocols and wireless channel properties, a network simulator is often used. In later development stages, the use of radios with full V2X software stacks is essential. This paper shows how an open integration and test platform enables model reuse and supports V2X application development and testing during all stages.


Mechanisms for resilience and robustness in vehicular networks

Muhammad Awais Khan
Institute of Telecommunications
The UrbanSense cooperative sensing infrastructure offers efficient communications between fixed data-collecting units (DCUs) and vehicles in a smart city. Different system technologies such as wave, wi-fi and cellular can be used to capture information from distant areas. Applications of research work can include monitoring meteorological conditions, noise, air quality and citizen behavior using video cameras. Other applications could be for data dissemination for semi-autonomous driving, adaptive cruise control, touristic information and warning systems such as lane-departure warning system, accident and traffic congestion messages.


Cooperative ITS pilot deployment in Hungary to support CAD

Adam Nagy
Traffic management engineer
Hungarian Public Road Non-Profit PLC
Future mobility requires more and more online information even from the road. The Hungarian Public Road company is committed to supporting this development. C-ITS deployment started in 2015 in Hungary to demonstrate the use of C-ITS to exchange data through wireless communication technologies between vehicles and infrastructure (V2I). The first C-ROADS - Harmonized C-ITS Specifications for Europe – Release 1.3 compliant RSU deployment took place in Q4 2018. Urban deployment is planned in the near future to provide testing possibilities in open-road conditions, which can be simulated and analyzed later on, for example within the Zala Automotive Proving Ground.

13:40 - 15:00


Room A Data Management, Storage and Cloud Technology
15:00 - 18:00


Modern hybrid storage solution for sensor data and HIL/SIL simulation

Dr Stefan Radtke
Technical director, EMEA
Data rates of modern sensors are constantly growing due to higher resolutions, multiple sensor types and an increasing number of sensors per car. A typical test car today produces 20-30TB per day. Considering a fleet of 10 cars and 200 recording days, the enormous amount of 60PB must be stored, processed and archived. Other requirements are global ingest and global access for annotation, processing and other process steps. In this session we’ll discuss a modern hybrid scale-out solution that uses cloud as well as data-center implementations to address these challenges.


Need speed? Automotive datalogging at 10GBs and more

Bernhard Kockoth
Advanced development lead
The main driver for high data rates in automated vehicles is growing sensor resolution for better environment understanding. Current setups require a number of high-resolution cameras and radar systems. Compression of data is traded for safety to achieve lowest latency from sensor to drive computers – the car can drive faster! OEMs and Tier 1s require all raw sensor data to train ML on the very same data the central drive computer 'saw' when the vehicle was on the road. We faced the challenges of 10 Gigabits and more and present a handy solution for data recording and transfer.


ADAS data gathering 2.0

Richard Levy
EU Platform UG
In the development of ADAS, data is generally collected in-vehicle and transferred to a low-cost site for labeling. The collection and transfer process wastes resources and time and creates risk. Data loses value quickly as sensors change. Sharing for fusion projects adds difficulty. Our test platform allows real-time annotation of raw data within the vehicle, and manual correction. We retain the ability to simulate multiple vehicle platforms with the same collected data. The cybersecurity implementation allows work with multiple partners on one vehicle, for example in fusion projects where camera, lidar and fusion components come from competitor companies.

16:15 - 16:45



Integrated data management for efficient validation of autonomous vehicles

Franz Gaber
Program Manager Data Management & Processing Platform
CMORE Automotive GmbH
Integrating various stakeholders, disciplines and locations into a data management system is one of the key challenges within the development and validation of autonomous vehicles. An intelligent workflow engine is required as well as integration of, and compatibility with, big data technologies. In addition, collaboration between huge teams, e.g. for data annotation and test, and management, must be supported. The presentation will show how CMORE Automotive's modular and highly flexible data management platform addresses these needs.


High-bandwidth datalogging solution for autonomous driving

Andreas Ehrle
X2E GmbH
With the new bandwidth demands arising from assistance systems and autonomous driving, a new architecture for datalogging devices is needed. The solution is a distributed datalogging system with a central datalogger and additional tap devices that can be located everywhere in the car. They send the data via an Ethernet link to the datalogger. For a high-speed infrastructure, 10GBit links between the datalogger and tap devices are needed. A distributed system like this has many advantages. Tap devices for different bus systems can be chosen and installed as needed for a specific car infrastructure.


Creating an optimized storage environment for autonomous vehicles

Jason Coari
Director, high-performance solutions
Quantum Storage
As organizations focus heavily on autonomous vehicle development, they are quickly learning that there is a massive volume of data rapidly growing in parallel to this activity – data that is extremely valuable and requires access in a time-sensitive manner. Within this context, it is imperative that the industry be better informed of data management solutions for this valuable data. This presentation will focus on how these organizations can create balanced storage architectures that can accelerate the ingest phase of this data from car to data center, and more cost-effectively preserve it over its lifecycle.

Room B Using Simulation for Test Scenario Validation
09:00 - 12:50


Testing and validating autonomous vehicles using traffic simulation

Dr Jochen Lohmiller
Manager microscopic simulation
PTV Group
Before testing connected and autonomous vehicles (CAVs) in real traffic on public test areas such as in Karlsruhe, Germany, virtual testing of those test areas using traffic simulation accelerates the development. Advanced traffic simulation models simulate all modes, such as individual human drivers, cyclists, pedestrians and different CAV behavior, which allows testing under different vehicle/driver populations. We present in a case study of how virtual tests can be applied using co-simulations between ego-vehicle simulation and a traffic simulation in the example of signal communication (V2X).


Massive simulation: recent achievements and perspectives

Thomas Nguyen That
Head of automotive domain
AV Simulation
Simulation has a key role to play in demonstrating the safety of autonomous vehicles, enabling billions of miles of virtual driving. This requires dedicated tools and methodologies, scenarios and models for the environment, sensors, and vehicles that are highly representative of the real world. AV Simulation is developing SCANeR studio, a versatile engine for automotive systems simulation. In this presentation we will present our latest experiences with massive simulation and explain how we addressed the challenges of model representativity, scenario generation, massive parallel execution and metrics computation.


Integration of simulation tools in validation processes

Ahmed Yousif
Software design engineer
Valeo Schalter und Sensoren GmbH
The presentation talks about the use of simulation tools to aid the validation of autonomous systems and algorithms. These systems can include lidars, cameras, ultrasonics and radar sensors. The use of simulation tools will help to reach the mileage required to validate the driving assistance systems in Level 4 and 5. In addition, the use of these simulation tools is needed where the artificial intelligence algorithms are exposed. The presentation will include some demos for lidars and ultrasonics-based algorithms such as full automatic parking.


The driver-in-the-loop simulator in the ADAS development and validation toolchain

Jelle van Doornik
Product manager ADAS
All car manufacturers are investing heavily in ADAS and AD development; billions of test kilometers are driven in simulation, and proving grounds are built specifically for the development and validation of (semi-)autonomous systems. A driver-in-the-loop simulator is effective to save time and cost, and speed up driver or passenger acceptance. Corner cases with different scenarios and conditions can be evaluated quickly. The challenge is to create an open architecture for engineers to work with in the development, and have standardized interfaces and content available to perform quick validation. This presentation explains how the driver-in-the-loop simulator fits in the total toolchain.

10:40 - 11:10



Machine learning and a systems approach to creating autonomous vehicle test cases

Siddartha Khastgir
Principal engineer
WMG, University of Warwick
To prove automated driving systems are safer than human drivers, it is suggested that they will need to be driven for over 11 billion miles. However, rather than the number of miles it is the quality of those miles that is important. To find the 'interesting test cases', a novel machine learning and systems-engineering-based approach has been developed to identify system failures, applied to low-speed automated driving systems (Pods). The approach also helps identify test cases from simulation environments that need to be carried over to real-world testing.


Real, robust and comprehensive – smart simulation for AV validation

Laksh Parthasarathy
Business head - automotive (autonomous validation)
Tata Consultancy Services
To ensure the safety of AVs, it is said that a vehicle needs to be validated by driving over 11 billion miles. However, beyond the number of miles that need to be racked up, it’s the miles where the vehicle encounters incidents that matter. Physical testing is not economically viable – we have to consider simulations. Some challenges with simulation includes identifying 'as many' edge cases as possible – where the software is likely to fail – along with an accurate representation of the world. In this paper, we explore unique techniques that could be leveraged to make simulations real, robust and comprehensive.


Physics-based simulation for ADAS and autonomous testing

Chris Hoyle
Technical director
Physics-based simulation allows us to model and simulate the real world and the sensors that feed perception systems. Taking a physics-based approach to simulation ensures that the results are not just convincing for humans, but also for machine vision systems, which require far greater levels of fidelity to deliver behavior that correlates with real-world testing. rFpro will demonstrate the techniques being adopted on UK Government- and EU-funded projects in which it is collaborating, aimed at delivering a roadmap to regulatory approval for Level 4 and Level 5 autonomous vehicles.


Vehicle-in-the-loop: validation of automated driving functions with virtual elements

Paul Prescher
Team manager
Future automated driving functions are supporting ever greater parts of driving scenarios in increasingly complex situations. This calls for new approaches to safe and resource-conserving testing and validation. To extend the test process to cover complex and safety-critical systems, such as automated driving functions, IAV has developed an approach that combines a real-world test vehicle with virtual elements for reproducible, safe and resource-saving tests. This opens up possibilities for validating automated driving functions both in the early as well as in the late phases of the development process.

12:50 - 14:00


Workshop - dSPACE

Further information to follow shortly. Check back soon for details.

Room B Using Simulation for Test Scenario Validation Cont...
16:00 - 18:10

16:00 - 16:30



Certification of autonomous vehicles in synthetic environments: initial findings

Timothy Coley
Product specialist
XPI Simulation
XPI Simulation, Warwick University and Thales are undertaking a feasibility study to examine the certification of autonomous vehicles in synthetic environments. The research aims of this study include: identifying resolution, rate and type of data required to represent the virtual world with a view to informing standardization activities; understanding the security vulnerability of a synthetic system being used in this manner; developing a certification approach that would achieve the expected repeatability and robustness within the synthetic environment to test autonomous vehicles. This presentation will cover initial findings from the research activity.


Development of model predictive motion planning and control for AD

Vishwas Sharma
Project Engineer, CAE
The presentation will discuss generation of optimal trajectory for autonomous vehicles, advanced control guidance for autonomous vehicles, and real-time validation in virtual environments. The aim of the paper is to present a system that can generate a real-time online optimal feasible trajectory to maximize comfort, stability and travel efficiency of autonomous vehicles while overcoming their physical constraints. In order to validate and tune the algorithms, several dynamics maneuvers and realistic urban driving scenarios are simulated. In the second half of 2019 it is planned to test the algorithms on a driving simulator.


Iterative real drive simulation in HILS

Sreeraj Arole
HiLS technical lead
Tata Elxsi
Vehicle simulation in a lab environment has played a key role in development and testing in the automotive industry. However, even when actual vehicle parameters and closest real-world scenarios are incorporated in the lab environment, there are several situations where a real vehicle is needed for testing/calibrating/recalibrating vehicle functionalities. This paper presents an effective way of using the network log from a vehicle combined with external sensor simulation and plant model running on a high-precision real-time simulator and control software to achieve a repeatable robust simulation environment. This enables a drive cycle with multiple variations to be undertaken to assess the performance and for calibration.


Open simulation interface as an enabler for virtual sensor models

Kmeid Saad
Scientific researcher
Hochschule Kempten
In the process of developing advanced driver assistance systems (ADAS), relying on virtual environments for the development process provides the advantage of entirely reproducible and controlled environment conditions. However, the existence of various function developers and simulation environments imposes the need for a generic and standardized interface. To tackle this issue, the open simulation interface (OSI) is created. OSI is a generic interface for the environment perception of automated driving functions in virtual scenarios. This paper demonstrates, with the help of OSI, the virtual implementation/simulation of a 'semi-physical camera model' and a 'behavioral multi-sensor model' in various simulation environments.
Please Note: This conference programme may be subject to change


Autonomous Vehicle International magazine