Reply Autonomous Days 2021

19 and 20 May 2021
Online Edition

Register now

AUTONOMOUS TECHNOLOGY

Autonomous vehicles such as cars, drones, robots and many more are becoming increasingly popular. They can be used in many different fields for daily or advanced and potentially dangerous tasks previously performed by humans. Fully free navigating vehicles are not wired and use a series of different sensory systems. This means they can give huge flexibility to industries across the board.

The new autonomy of things is the direct result of technological advances in various fields such as AI and Edge-to-Cloud Computing. Thanks to computer vision, object and pattern recognition autonomous vehicles are able to find their path while constantly collecting data with equipped sensors in any form, from cameras to microphones, GPS, thermometers, moisture sensors, gas or radiation detectors. Wireless connectivity with an intelligent system architecture and Cloud-Edge Computing allows for the leveraging of collected data with AI & Machine Learning algorithms, enabling the device to take its own decisions on how to best perform a specific task.

The value chain of data

Join the Reply Autonomous Days and be invited to take a look along the data value chain. Together we will explore, which topics need to be considered when developing autonomous systems.

Let's dive into the different areas and implementation possibilities of autonomous systems during the two Autonomous Days 2021.

Agenda Day 1

3:00 pm - 5:30 pm - 19 May 2021 - Online Edition

Download the booklet for the event.


Mobility
Perception
Cognition
Security
ROS
3:00 pm - 3:20 pm
Welcome
Introduction to Autonomous Things
Everyone is familiar with the development spiral in the automotive sector with its ever increasing number of functions. It is always about assistance functions coupled with connected services.
The path towards self-driving vehicles is clear. A similar movement comes from the industry sector using robots, which take on more and more functions independently. These introduction provides an overview on the current trends and challenges.
Peter Schiekofer
Autonomous Reply
3:20 pm - 3:40 pm
Mobility
Smart City mobility concepts in Regensburg
Focus on autonomous shuttle vehicles as part of the mobility of the future
Within the Smart City context, the mobility sector is one of the fastest growing segments. Smart mobility is characterized by digital processes, transparency, multimodal transport offers and on-demand services.
Small autonomous shuttle vehicles are part of this future. With the pilot project “Autonomous Shuttle Regensburg”, das Stadtwerk Regensburg. Mobilität would like to lay the foundation for the broader use of autonomous shuttle vehicles in public transport systems
Marco Krakowitzer
Das Stadtwerk Regensburg.Mobilität GmbH
3:40 pm - 4:00 pm
Perception
LiDAR sensor technology for autonomous systems
Recognition Software for Real-World Applications
In his talk, Florian Petit will give an introduction to LiDAR technology and 3D sensing. It is demonstrated how the wealth of captured data can be leveraged using recognition software to enable real world applications such as people counting, traffic management, and automated and autonomous driving.
Florian Petit
Blickfeld
4:00 pm - 4:20 pm
Cognition
Small Data is the new Big Data
Current Challenges in Perception
The rise of Deep Learning has led to huge improvements in perception systems. Without these perception systems based on Deep Learning, the development of autonomous things like cars or drones for real-world applications with high demands on accuracy and reliability would not be possible. Nevertheless, current systems require large amounts of training data and are by far inferior to human perception in many aspects. Examples of these aspects are generalization performance and vulnerability to adversarial attacks. In this talk, the current challenges of Deep Learning based perception systems will be presented, a short overview of research activities to overcome these challenges will be shown and an outlook on what is needed in general to outperform human perception will be given.
Torsten Hafer
Autonomous Reply
4:20 pm - 4:40 pm
Security
Cybersecurity in the driving seat with ISO21434
The cybersecurity management system standard which drives the entire life cycle of connected vehicles
Digitalization is everywhere and connecting items is necessary in the world of cars - now also called smart cars. “Smart” means a lot of code and data. This creates a larger attack surface, and the risk of hacker attacks increases. Cybersecurity as the unequal twin of safety needs to be taken care of. For this there is a guidance in the form of a Cybersecurity management standard for road vehicles called ISO 21434.
It provides a cybersecurity framework for product developers, OEM’s, suppliers, or any other participants in the supply chain for the entire life cycle of road vehicles, from development to decommissioning and includes risk assessment and management methods. This enables to secure by design and in operation and applies to all electronic systems, software, and connectivity.
This session will briefly introduce the ISO 21434 standard and explain the key building blocks, discuss challenges in operationalizing and give examples of how specific areas can be implemented.
Ibrahim Köse
Spike Reply
4:40 pm - 5:00 pm
ROS
Robotic Operating System (ROS): From a university project to a universal standard.
The importance of a ROS for Autonomous Systems
In this talk we will cover the use of ROS for the development of Autonomous Systems. We will start with a brief introduction to ROS, tracking its evolution from its beginnings at Stanford to its modern-day race to become a universal standard. A short discussion of its strengths and weaknesses will be presented followed by an overview of its versatility for multiple applications in the fields of Perception, Mapping and Localization, Navigation and Simulation.
We will cover the main improvements of ROS2 in comparison to its predecessor ROS1 and look at possible applications where these improvements will be beneficial, as well as at real-life examples, where these improvements have been proved to work. The talk will end with a short video-recorded demonstration that highlights the ease and versatility of development in ROS.
Ricardo Cedillo
Autonomous Reply
5:00 pm - 5:30 pm
Q&A
Closing of the day

Agenda Day 2

3:00 pm - 5:30 pm - 20 May 2021 - Online Edition

Download the booklet for the event.


Organization
Data Ingest
Virtual Testing
Machine Learning
Virtual Simulation
3:00 pm - 3:10 pm
Welcome and Summary of Day 1
3:10 pm - 3:30 pm
Organization
Rollout Management for Firmware Over the Air (FOTA) Updates
Cloud Solutions for valid Software Configurations
In this session, we will introduce you to the architecture of the digital truck platform at Linde Material Handling (LMH) - one of the leading manufacturers of forklifts worldwide - and show how Firmware Over the Air (FOTA) updates are rolled out to their customer fleets.
With the implementation of digital twins, LMH launched a new generation of connected forklifts. This new generation comes with enhanced connectivity and a revolutionary electrical architecture that is highly efficient and safe. It includes improved visibility, safety assistance systems and new software features, which can be purchased and activated ‘Over the Air’. To make this happen, we jointly developed an easy-to-use cloud solution that can reliably ensure valid software configurations and trigger and control software updates remotely.
Benjamin Schürmann
Concept Reply
3:30 pm - 3:50 pm
Data Ingest
Event based recording of raw data in ADAS/AD systems
Efficient acquisition and selection of valuable data for automated driving
Petabytes of raw data from high-resolution sensors form the essential basis for function development, but also represent a major hurdle in the development process. The presentation deals with the efficient acquisition and selection of valuable data for automated driving and thus separates the wheat from the chaff.
In this talk we will
  • give Engineers a technical overview of a modular measurement architecture which covers the needs for processing the high data rates in the vehicle for sensorics like Camera, Radar and LIDAR
  • give Algorithm developers insights how to filter corner cases already in the vehicle
  • give Technical management a recommendation how to increase efficiency in test fleets
Alexander Noack
b-plus Technologies GmbH
3:50 pm - 4:10 pm
Virtual Testing
Autonomous Driving Vehicles Validation
Challenges and Solutions
Autonomous Driving Vehicles are hitting the road today. During the last years, the complexity of these systems increased dramatically and, to ensure acceptable level of safety, efficient E2E verification and validation solutions are necessary. One of the main challenges derives from the need of managing a large number of tests and scenario variations. Concept Reply developed in-house methodologies to support customers during the difficult journey to successfully perform requirements-based and scenario-based virtual testing activities.
Simone Maragliulo
Concept Quality Reply
4:10 pm - 4:40 pm
Machine Learning
Finding Parking Spots
Machine Learning, Routing & Personalization
In this talk we will demonstrate how to guide drivers on a “park search route” which optimizes
  • the probability to find an on-street parking spot
  • the walking time from parking spot to destination.
We will show how machine learning is used to find out probabilities for free parking spots and which data is needed to train a model.
Additionally a new kind of routing algorithm is demonstrated which optimizes search and walking time and supports a driver’s preferences.
Lars Haferkamp
Comsysto Reply
4:40 pm - 5:00 pm
Virtual Simulation
What is the relevance of virtual simulations for the development of future autonomous driving systems?
A simulation-based approach
To ensure the safety of autonomous driving functions in all conceivable traffic situations, it is essential to consider an extremely large number of scenarios in development and testing. A decisive argument in favor of simulation is the lack of possibility to detect all scenario variations in real driving tests or to reconstruct them on the proving ground. A simulation-based approach allows for flexible and scalable testing of autonomous vehicles in the virtual world.
Marina Liebich
IPG Automotive GmbH
5:00 pm - 5:20 pm
Final discussion
5:20 pm - 5:30 pm
Closing of the Day
Peter Schiekofer
Autonomous Reply

Speakers

Peter
Schiekofer

Autonomous Reply

Marco
Krakowitzer

Das Stadtwerk Regensburg.Mobilität GmbH

Florian
Petit

Blickfeld

Torsten
Hafer

Autonomous Reply

Ibrahim
Köse

Spike Reply

Ricardo
Cedillo

Autonomous Reply

Benjamin
Schürmann

Concept Reply

Alexander
Noack

b-plus Technologies GmbH

Simone
Maragliulo

Concept Quality Reply

Lars
Haferkamp

Comsysto Reply

Marina
Liebich

IPG Automotive GmbH

Get in touch

Save your seat now for the Autonomous Days 2021.
Register now