Software Development for Autonomous Driving
22 - 24 October 2019
Berlin, Germany
Agenda Day 2
8:30 am - 9:00 am
Registration and welcome coffee
9:00 am - 9:10 am
Opening remarks
Prof. Dr. Stefan Leue -
Chair of Software and Systems Engineering,
University of Konstanz
9:10 am - 9:50 am
Reducing lead time and increasing quality: simulation based verification of autonomous and classic software
Henrik Lönn -
Technology Specialist, Embedded SW Development,
Volvo
Simulation based verification is a means to increase quality and reduce lead time in automotive
software development. It is particularly important for autonomous vehicles, as many of the test
cases would be too rare or dangerous to perform with a real vehicle.
This talk will present an approach to represent, generate and execute simulations, covering the
range from unit testing of software components up to the complete vehicle.
9:50 am - 10:30 am
From model based to learning based control architectures
Thomas Raste -
Head of Global Chassis Control,
Continental, Germany
For lower dimensional software in control applications is more easy to use structural analysis and
quality assurance for, when it comes to AI systems. We will discuss the role of motion control in
autonomy stack and discover the journey from model based to learning based control applications,
using neural networks.
10:30 am - 11:00 am
Refreshment break and networking
11:00 am - 11:40 am
Safe but not cautious path planning: combining conventional methods with machine learning
Mattias Brännström -
Technical Expert Precautionary Safety,
Zenuity
• What are the safety requirements for AD?
• What does it take to close the safety gap between ADAS & AD?
• Addressing the challenges by combining conventional methods with machine learning
11:40 am - 12:20 pm
Architecture centric approaches in engineering of autonomous software for improved quality and development time
Pablo Antonino -
Department Head „Embedded Software Engineering“,
Fraunhofer IESE
Evaluating various architectures for robustness and operability with precision is always a challenge
in embedded systems. This presentation offers a study on the ways it has been done across
various automotive companies and derives results for future processes. It also outlines
cross-enterprise issues such as team management for quality architectures and offers solutions
in this area.
12:20 pm - 1:50 pm
Networking luncheon
1:50 pm - 2:30 pm
Round table sessions
Prof. Dr. Stefan Leue -
Chair of Software and Systems Engineering,
University of Konstanz
Henrik Lönn - Technology Specialist, Embedded SW Development, Volvo
Round table sessions | Choose your main discussion topic and deepen your knowledge in close
dialogue with experts by pointed questions.
Table 1 | How to combine MBSE and Agile in safety development
Stefan Leue, Chair for Software and Systems Engineering Department, University of Konstanz
Table 2 | Standardizing the overall software architecture for autonomous systems
Henrik Lönn, Senior Principal Research Engineer, Volvo Group Trucks Technology
2:20 pm - 3:00 pm
Software Architectures and ASIL Decomposition Across AI and Open Source Technologies
Riccardo Cagnacci -
Functional Safety Senior Architect,
Intel
Open source powers the software-defined infrastructure that transformed the modern data center
and ushered in the data-centric era. Today, the vast majority of the public cloud runs on open
source software; new contributions by Intel are poised to drive a future where everything is
software-defined, including new areas such as automotive, industrial and retail. We will discuss the
peculiarities of decompositions for ASIL levels.
3:00 pm - 3:40 pm
Precise and Robust Vehicle localization to Handle Complex Highway and Urban Use-Cases
Benazouz Bradai -
Autonomous Driving Innovation Platform,
Valeo
This talk will decribe precise and robust crowd-sourced simultaneous localization and mapping
solution (SLAM), which enables centimetric accuracy using onboard Lidar sensors, affordable
automotive grade IMU and low cost GNSS receiver. It works in poor and denied GPS conditions
(tunnel, multipath issue...), it is suitable for indoor/outdoor parking and large scale areas and can
be complementary to existing HD Maps.
