www.auto-ai.com Open in urlscan Pro
18.195.249.37  Public Scan

Form analysis
0 forms found in the DOM

Text Content

book now

Auto.AI USA | June 19 – 21, 2022, The Henry Hotel, Detroit
EVENT APP LOGIN
 * Home
 * Agenda
   * View Web-Agenda 2022See all the exciting sessions we have in store for you
   * Download Agenda
   * BOOK NOW
 * Speakers
   * Panel 2022Find out which experts will share their insights and use cases at
     Auto.AI USA
   * Speaker CommunityFind out which experts shared their insights and use cases
     on previous Auto.AI USA editions
   * Become a Speaker
   * BOOK NOW
 * Partners
   * Business Partners 2022Network with our partner community and discover their
     cutting-edge products & services
   * Why Partner?Connect with the right people in a space designed to optimize
     your relationship-building and deal-making
   * Matchmaking & One2OnesMeet the right decision makers based on your interest
   * Get Partner Packages
   * Get Attendee List 2022
   * Get Attendee List 2021
 * Who & Why?
   * The Event at a glanceWhat can you expect as an attendee at Auto.AI USA
   * AttendeesYou will meet decision makers from the industries that matter to
     you – digital or live!
   * 10 reasons to attendWhy you should join
   * Get Attendee List 2022
   * Get Attendee List 2021
 * Event Info
   * Travel & LocationOnsite Location - Join us in Detroit!
   * Evening EventsIcebreaker + Networking Dinner
   * Co-located EventOSS.5 USA - No. 1 event on functional, system & operational
     safety for highly to fully automated vehicles
   * Review 2021Everything gathered about the last years' event
   * ContactGET IN TOUCH - We are here for you!
   * BOOK NOW
 * GET AGENDA
 * BOOK NOW

Multi-Touchpoint Networking our event agenda THE
AGENDA
The most effective, innovative and informal networking hub of the year!


AGENDA



Loading...

Event days

Select all

19.06.2022

ICEBREAKER



20.06.2022

EVENT DAY 1



21.06.2022

EVENT DAY 2




Categories

Select all

AI + MACHINE LEARNING + DEEP LEARNING

CONNECTIVITY + V2X

DATA

NETWORKING

ODD + SAFETY + SECURITY

PERCEPTION TECH + SENSOR SYSTEMS

PRESENTATION

SENSOR FUSION + COMPUTER VISION

SOFTWARE ENGINEERING

TRAJECTORY PLANNING + PREDICTION + DECISION-MAKING

VALIDATION + SIMULATION + MODELING

WORKSHOP



Locations

Select all

Dinner Location

Exhibition Area

Icebreaker Location

Plenum

The Henry, Autograph Collection

Virtual Meeting



Companies

Select all

AEye

Algolux Inc.

Amazon Web Services (AWS)

Ambarella

Aurora Inc.

Changan Automobile Co Ltd

CloudFactory

Cruise Automation

Daimler Trucks North America

Evonomy

Ford Motor Company

Fraunhofer IKS

General Motors

Hyperspec AI

Innovusion

Locomation

May Mobility

Motional

Plus

Pony.Ai

rabbitAI

Ridecell Inc.

Run.ai

Sama Inc.

Toyota North America

Waymo

we.CONECT Global Leaders GmbH

Weights & Biases

Workshop
Icebreaker Sessions – Break the ice and get the show on the road!
19.06.2022
18:45 - 22:00 (GMT-05:00) Eastern Time (US & Canada)
Icebreaker Location

What is the Icebreaker Session?



The Icebreaker session takes place on the evening before the event starts.

Just when everyone arrives, we are inviting you to join us for some early
networking and give you the chance to settle in the atmosphere of the event.
This is the time and place to share your thought-provoking ideas and get things
off an interesting start. Glass in hand, make your way around the round tables,
discuss and network with the speakers and business partners. We invite you to a
unique location and allow you to get to know your fellow conferences attendees
in a relaxed and interactive setting.

 * How is it organised?In order to launch the conference, relevant topics and
   questions will be touched on moderators at their round tables.
 * Without slides, moderators will present their project or thoughts on a
   particular approach and enter into an open discussion with interested
   participants.
 * During the Icebreaker Session, participants can move freely between the round
   tables and get involved in conversation.
 * Discussions evolve, either led by the moderator or on a personal basis
   between individual participants.

Break the ice and get the show on the road!


Ai + Machine Learning + Deep Learning Workshop
1 I Icebreaker - Are we Currently Unleashing the Full Power of AI, Machine
Learning and Deep Learning in the Development of Autonomous Vehicles?
19.06.2022
18:45 - 22:00 (GMT-05:00) Eastern Time (US & Canada)
Icebreaker Location
Robert BloomquistVice President of Automotive Business Development
Ambarella
 * What common applications of AI, ML, DL outside of the automotive industry
   should we also be exploring?
 * Why deep learning alone won’t give us level 5 self-driving cars?
 * How intelligent really are our AI, Machine Learning and Deep Learning
   approaches?


Odd + Safety + Security Software Engineering Validation + Simulation + Modeling
Workshop
2 I Icebreaker - What Are the Hardware Technologies that are Likely to Influence
Adoption of Self Driving Systems?
19.06.2022
18:45 - 22:00 (GMT-05:00) Eastern Time (US & Canada)
Icebreaker Location
Gaurav SinghSystem Architect
Ridecell Inc.
 * What Are the Hardware Technologies that are Likely to Influence Adoption of
   Self Driving Systems? E.g. Neuromorphic computing hardware which promises to
   increase compute efficiency and power requirements or high definition radars.
   What are the key challenges for such technologies and what is been done to
   overcome them?
 * How to efficiently scale self driving cars to different cities and
   geographies with unique driving styles – is collecting more data and
   scenarios from human driven fleets the way?
 * Why is there no standardization in ODD definition and key metrics in self
   driving car industry? Will there be advantages in pooling challenging test
   cases and following standards?
 * Will vision transformer deep learning architectures replace CNN and LSTM
   style architectures for perception and prediction tasks ? Can the on-board
   processing hardware support such architectures?




Data Perception Tech + Sensor Systems Validation + Simulation + Modeling
Workshop
3 I Icebreaker - What are the challenges of using real world data augmentations
for perception tasks?
19.06.2022
18:45 - 22:00 (GMT-05:00) Eastern Time (US & Canada)
Icebreaker Location
 * How should we be using data to improve sensor performance and robustness?
 * What are the challenges of validation and testing of the data-driven AI
   functions?
 * What are the main constraints, areas and strategies to scale up AI-based
   autonomous drivingThis is an interactive session that requires no slides.


Networking
REGISTRATION & AUTO.AI COMMUNITY WALL
20.06.2022
07:45 - 08:30 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection

The official registration for the conference: Pick up your name batch, take a
Polaroid photo for the Auto.AI Community Wall, use your personalized matchmaking
app, make first contacts & plan meetings and get started. Welcome to the Auto.AI
USA!


Presentation
Welcome & Introduction by the Program Director & the Conference Chair
20.06.2022
08:45 - 08:55 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Sarah FarleyPortfolio Director Smart Mobility & Automotive
we.CONECT Global Leaders GmbH

Networking
Say Hello!
20.06.2022
08:55 - 09:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum

Get the show on the road! To warm up for the event we invite you to share your
expecta tions in the event chat & discover who is joining the conference.


Ai + Machine Learning + Deep Learning Data Odd + Safety + Security Perception
Tech + Sensor Systems Presentation
Keynote I Key AI and Machine Learning Challenges in Development of L4+
Autonomous Vehicles
20.06.2022
09:00 - 09:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Sammy OmariVP Engineering, Head of Autonomy
Motional

Mobility is a core human need, like eating and drinking. Solving L4+ driving is
the single most important engineering challenge of our time. It has the
potential to dramatically change the world by reducing accidents, carbon
emissions, and parking lots, but it will involve solving some of the most
difficult AI and Machine Learning problems we have ever encountered. Join me
while we explore some of the key AI and ML challenges involved in developing L4+
autonomous vehicles, such as dealing with noisy sensors, finding the data
needles in the haystack, processing multi modal data in concert, leveraging
unlabeled data, reasoning about unknowns, and much more.

 * Handling sensor noise and degradation is one of the biggest challenges for AV
   Perception.
 * Novel methods are emerging for leveraging the treasure trove of unlabeled
   data collected by AVs.
 * Due to the safety-critical nature of AV technology, we must blend
   Machine-learned approaches with normative decision making.
 * Predicting human behavior is an unsolved problem but new technology is
   emerging that can leverage a fleet of human-driven cars to aid in learning
   this
 * task.


Data Presentation
Solution Study I Lane Detection for AVs - Distributed Training at Scale
20.06.2022
09:30 - 10:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Jonathan CosmeAI/ML Solutions Architect
Run.ai

The deep learning models driving innovation in autonomous vehicles are becoming
more ambitious by the day, but their supporting infrastructures often struggle
to keep up. Because a single GPU can’t accommodate the complex neural networks
of enterprise AV projects, distributed training has emerged as the solution for
training DL models and large data sets. In distributed training, storage,
compute power and batch size are magnified with each GPU added to the cluster,
dramatically reducing training time.

 

In this talk, we address a lane detection use case where Run:ai, Microsoft, and
NetApp jointly built a distributed training DL solution at scale that runs in
the Azure cloud. This solution enables data scientists to fully embrace the
Azure cloud scaling capabilities and cost benefits for automotive use cases.

 

You’ll learn:

 * The compute challenges you can expect to face as your teams and models scale
 * How we managed to keep tens of GPUs fully occupied at 95% to 100% utilization
 * The role of resource allocation in accelerating model training in the cloud


Presentation Trajectory Planning + Prediction + Decision-making
Case Study I Autonomous Driving Using Multi-Policy Decision Making (MPDM)
20.06.2022
10:00 - 10:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Kamil LitmanVice President of Software
May Mobility

In this talk, I’ll describe an approach to building autonomous vehicles using
Multi-Policy Decision Making (MPDM). MPDM is a method for high-level decision
making, in which a vehicle makes decisions like “should I pass this bicycle, or
should I follow it?” It enables autonomous vehicles to interact more naturally
with other cars and road users like bicycles and pedestrians, which is a
requirement for autonomous to be widely deployed. I will also describe our
ongoing commercialization of self-driving technologies (including MPDM), which
we believe can help to transform cities and increase transportation access and
equity.

 * Building behaviors that generalize across a wide range of scenarios is
   essential to autonomous driving
 * Rule-based systems become cumbersome and costly to maintain due to an endless
   edge cases
 * MPDM is May’s approach to building behavioral planners that can handle new
   and unexpected scenarios
 * MPDM has been deployed in mixed traffic on public streets across seven cities


Networking
BREAKFAST BREAK & NETWORKING
20.06.2022
10:30 - 11:00 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Get coffee, get snacks, make connections and talk with your peers – and have the
chance to actively participate in product and service demos on the expo floor.


Networking
100 MINUTES – ONE2ONE MEETING SESSIONS (01 | 02)
20.06.2022
10:30 - 11:00 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Through our AI-based matchmaking technology we identify and determine interests,
consultation requirements, consulting expertise and solutions as well as
currently ongoing projects of the participants. This assessment allows for
qualified and time-efficient One2Ones with your colleagues and/or potential
partners. Each session lasts about 10 minutes.

Start of Session 01 & 02 – Search – Find – Match – Network


Presentation Trajectory Planning + Prediction + Decision-making Validation +
Simulation + Modeling
Case Study I End-to-end Joint Object Detection and Motion Forecasting Models for
Autonomous Driving
20.06.2022
11:00 - 11:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Nemanja DjuricStaff Autonomy Engineer, Machine Learning Tech Lead
Aurora Inc.

Object detection and motion forecasting are critical components of self-driving
technology, tasked with understanding the current state of the world and
estimating how it will evolve in the near future. In the talk we focus on these
important problems and discuss end-to-end methods that jointly perform these two
tasks, which have shown state-of-the-art performance. We present a deep dive
into recently proposed MultiXNet end-to-end model, and discuss a number of
improvements that build on this method as a base.

 * learn about the problems of object detection and motion forecasting
 * become familiar with the state-of-the-art approaches that jointly perform
   these two tasks
 * learn about various modeling improvements that resulted in significant boosts
   in performance


Perception Tech + Sensor Systems Presentation
Solution Study I Long-Range Image-Grade LiDAR for Autonomous Vehicles from
Innovusion
20.06.2022
11:30 - 12:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Erich SmidtManaging Director
Innovusion

As a worldwide leader in the design and development of long-range image-grade
LiDAR, Innovusion will present and demonstrate the complete portfolio of their
LiDAR system. Unprecedented insight into the environment far in front of the
car, with a long-distance range detection while simultaneously achieving
ultra-high resolution. Innovusion long-range image-grade LiDAR can:

 * detect objects with 10% reflectivity at a distance of 200 meters, creating
   conditions for intelligent networked vehicles to drive at high speed.



At the same time, the angular resolution of 0.10° enables enables easy detection
of objects measuring 20x20x20cm at a distance of 120 meters thereby allowing a
sufficiently safe braking distance for the intelligent networked car. With a
product design that meets the requirements of vehicle regulations, it highlights
high performance while ensuring stability and long life cycle. Falcon is a good
choice for intelligent networked automotive LiDAR sensors of L3 and above.


Data Presentation Trajectory Planning + Prediction + Decision-making Validation
+ Simulation + Modeling
Case Study I Challenges in conceptualizing ADAS simulation models – the issue of
motion recognition of other vehicles on urban intersections
20.06.2022
12:00 - 12:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Honghao TanPrincipal Engineer
Changan Automobile Co Ltd

To simulate crossing traffic at intersections and on complex road networks,
vehicles must traverse intersections with a specified behavior. The vehicle must
be able to detect other vehicles and objects on and beside the road. In this
session we will cover what simulation models are best suited for ADAS to deal
with everyday scenarios in urban intersections.

 *  What are the current processes used to validate ADAS systems?
 *  Simulation is critical but how good is “good enough”?
 *  How to program ADAS to counterpredict the behavior of other vehicles
 * How much data is required to develop reliable prediction models?


Data Presentation Software Engineering
Solution Study I Clear, Clean, Composable - ML Pipelines for AV
20.06.2022
12:30 - 13:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Andrew TruongMachine Learning Engineer
Weights & Biases

The amount of experiments conducted during L4, L5 R&D is massive. How do you
manage such diverse set of experiments, each with different dataset,
architecture, and hyperparameter choices? How can you systematically capture and
version the data and code that makes up your pipelines, and make them available
for everyone on your team to analyze and learn from your latest experiment?

 * This session covers tooling to setup clean experimentation pipelines and
   share results across larger groups.


Networking
LUNCH BREAK & NETWORKING
20.06.2022
13:00 - 14:00 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Have lunch and make connections and talk with your peers – and have the chance
to actively participate in product and service demos on the expo floor


Networking
100 MINUTES – ONE2ONE MEETING SESSIONS (03 | 04)
20.06.2022
13:00 - 14:00 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection

Through our AI-based matchmaking technology, we identify and determine
interests, consultation requirements, consulting expertise and solutions as well
as currently ongoing projects of the participants. This assessment allows for
qualified and time-efficient One2Ones with your colleagues and/or potential
partners. Each session lasts about 10 minutes.

Start of Session 03 & 04 – Search – Find – Match – Network


Ai + Machine Learning + Deep Learning Data Presentation
Case Study I Ensuring Quality Data for Deep Learning in Varied Application
Domains
20.06.2022
14:00 - 14:30 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection
Data Collection, Curation, Annotation, and Analysis
Gaurav SinghSystem Architect
Ridecell Inc.

The presentation explores the data lifecycle for deep learning with a special
focus on data curation, quality annotations and data analysis in the context of
autonomous vehicles. For optimizing data curation, techniques such as active
learning will be explained and we will address the question how to choose which
data to send for annotation. Additionally, the session will discuss how to
select an annotation partner and how to efficiently do annotation in-house. And
we will highlight how to frame good annotation instructions for different
annotation tasks like lidar annotation, semantic segmentation and sequence
annotations. Finally, the session will provide solutions to common problems seen
in the curation and annotation process and how to overcome them.



 * How an optimal data lifecycle for deep learning should look like
 * How to choose the right data for annotation
 * How to frame annotation instructions for different areas (Lidar, semantic
   segmentation, sequence annotation etc.)
 * How to master the main challenges of your annotation process


Data Perception Tech + Sensor Systems Presentation
Solution Study I Building a Robust Automotive LIDAR Annotation Quality Rubric
20.06.2022
14:30 - 15:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Renata WalkerSenior Product Manager
Sama Inc.

In this talk, Renata will discuss:

 * The importance of building a robust automotive LiDAR Annotation Quality
   Rubric
 * Defining quality scoring and considerations for trade offs
 * Best practices for selecting an annotation partner who can deliver against
   your data objectives


Perception Tech + Sensor Systems Presentation Sensor Fusion + Computer Vision
Software Engineering
Case Study I Development of 8MPX automotive camera system
20.06.2022
15:00 - 15:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Timofey UvarovCamera System Lead
Pony.Ai

High-quality cameras are fundamental sensors in assisted and autonomous driving.
In particular, long-range forward-facing cameras can provide vital information
about the road ahead, including detection and recognition of objects and early
hazard warning. These automotive cameras should provide high-resolution images
consistently under extreme operating conditions of the car for robust operation.

The main focus of the presentation is the steps of designing an automotive
camera system based on an 8mpx sensor. We will cover key aspects of image
quality, system architecture, and benchmarking of 8mpx sensor using the same
size 2mpx sensor as a reference point.

 * How does pixel size and count determine the optical diagonal of camera
   sensors?
 * Does pixel size enable improved computer vision?
 * How different camera models cope with difficult environmental conditions?
 * What are the implications on iamge processing with the increase of pixel
   size?


Perception Tech + Sensor Systems Presentation Software Engineering
Solution Study I The Role of Dynamic Modulation and Sparsity in High Performance
Imaging Radar
20.06.2022
15:30 - 16:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Paul DentelSenior Technical Product Manager
Ambarella

Today, the state-of-the-art solution to enable high performance imaging radar is
to increase the antenna count and processing capabilities of the system.
Fundamentally, this approach is not an optimal solution for L4 applications as
the imaging radar available on the market use the same principles as basic
automotive radar. In contrast, utilizing dynamic modulation and sparsity in
multiple dimensions can create a significantly higher-performance radar system.
This approach enables performance to be scaled, not by the antenna count, but
rather by the compute capabilities of the radar system, which yields a more
optimal and higher-performance solution.



4 Key Takeaways:

1.      There are tradeoffs associated with state-of-the-art imaging radar
systems.

2.      Dynamic modulation can be used to enable high Doppler resolution and
range resolution.

3.      Sparsity in both hardware and software can be leveraged to achieve
better angular resolution.

4.      These approaches enhance performance and have wide applicability in both
edge and centrally processed architectures.


Networking
100 MINUTES – ONE2ONE MEETING SESSIONS (05 | 06)
20.06.2022
16:00 - 16:30 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection

Through our AI-based matchmaking technology we identify and determine interests,
consultation requirements, consulting expertise and solutions as well as
currently ongoing projects of the participants. This assessment allows for
qualified and time-efficient One2Ones with your colleagues and / or potential
partners. Each session lasts about 10 minutes.

Start of Session 05 & 06 – Search – Find – Match – Network


Networking
Afternoon Break & Networking
20.06.2022
16:00 - 16:30 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Get coffee, get snacks, make connections and talk with your peers – and have the
chance to actively participate in product and service demos on the expo floor.


Workshop
Challenge your Peers Sessions
20.06.2022
16:30 - 17:15 (GMT-05:00) Eastern Time (US & Canada)
Plenum

What is Challenge your Peers?

Challenge Your Peers is an innovative workshop concept which enables
participants to engage in a direct exchange with experts from a certain
industry. The aim is to discuss and to identify challenges within the sector as
well as problems, needs and solutions. In the run-up of the conference we.CONECT
gathers information about the main questions and interests of the participants.
During the session, they then discuss their specific issues and topics with
their peers on a mindmap designed by we.CONECT! Please choose a table according
to your interests and main challenges and discuss this – with your peers!

Challenge your Peers Session in
Pictures!https://www.youtube.com/watch?v=Vi9dTKb5-2E


Ai + Machine Learning + Deep Learning Odd + Safety + Security Software
Engineering Trajectory Planning + Prediction + Decision-making Validation +
Simulation + Modeling Workshop
1 I Challenge Your Peers - What Are the Requirements for the Modelling of the
Systems Involving Machine Learning Components?
20.06.2022
16:30 - 17:15 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Dr. Chih-Hong ChengDepartment Head - Safety Assurance for AI
Fraunhofer IKS
 * End-to-End solution for vehicle guidance: hype or a solution?
 * The requirements of machine learning/Deep-Learning algorithm for safe
   automated vehicle guidance?
 * Beyond the detection: The use of -Deep-Learning in order to predict the
   behavior of the road participants.
 * Deep-Learning based decision making: how does the architecture look like?
 * How to integrate the safety/redundancy to Deep Learning algorithm?


Connectivity + V2x Odd + Safety + Security Perception Tech + Sensor Systems
Sensor Fusion + Computer Vision Workshop
2 I Challenge Your Peers - Benchmarking Sensing & Perception – How can we
objectively evaluate alternative Solutions for Autonomous Vehicles?
20.06.2022
16:30 - 17:15 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Mike BucalaTechnical Fellow
Daimler Trucks North America
 * Must the ODD be limited to roadways with available HD maps, or can sensor
   fusion provide for a safe MRM?
 * When would “platooning” make sense in context of SAE Level 4 operation?
 * Why is Safety of the Intended Functionality particularly important for
   automated vehicle systems?
 * Are telematics and V2V communications able to interact with the autonomously
   operated vehicle in a way that ensures cybersecurity?
 * Can the ADS be expected to correctly perceive and react to first responders
   directing traffic outside the defined corridor of operation?


Data Perception Tech + Sensor Systems Sensor Fusion + Computer Vision Software
Engineering Workshop
3 I Challenge Your Peers - What are the Challenges and Perspectives of Edge
Computing as an Alternative to the Centralized Architectures for L4 Autonomous
Vehicles?
20.06.2022
16:30 - 17:15 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Vineeth VenkiteswaranGlobal Business Development Leader, Autonomous Vehicles
Amazon Web Services (AWS)

 * Differentiation between centralized and distributed compute systems: What
   system do we call centralized and what do we call distributed systems in
   automotive? What type of system is human vision?
 * Now-a-day trends and challenges in architecture: Existing systems lean to
   centralized architecture, though sensors are getting smarter and one can
   deploy a trained network on an ISP or sensor chip. What are the challenges to
   use such chips in L4 automotive pipeline considering:
 * • Cost /Computational power
 * • Power consumption / Infrastructure / Bandwidth
 * • Management of the system
 * • Synchronization and fusion of data from multiple sensors
 * What could be the example of a system / product using hybrid architecture?
   For example, a smart camera could use an internal onboard processor for
   better understanding of scene contents and improving the image quality.
 * What are the possibilities of future architectures? And how would the
   business landscape change accordingly?


Ai + Machine Learning + Deep Learning Data Odd + Safety + Security Presentation
Sensor Fusion + Computer Vision
Case Study I Why machine learning is unreliable? Assessing the safety issues in
autonomous driving
20.06.2022
17:15 - 17:45 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Tekin MericliCTO & Co-Founder
Locomation

A lot of automotive OEMs and Transportation companies emphasize the role of deep
learning in fostering autonomous driving, but can an immense set of data without
control be considered “safe” for the development of autonomous driving? What are
the pitfalls and drawbacks of deep learning? In this session we will discuss
what risks contemporary computer vision methods bring with them and take a look
on how traditional methods can help to make autonomous driving safer.

 * What are safety related concerns by machine learning?
 * What are the data standards in the automotive industry? Can we expect that
   the vehicle behaves in accordance with the data input?
 * Is end-to-end machine learning truly safe and necessary?
 * How can traditional computer vision methods assist in developing safer
   autonomous driving experiences?


Presentation
End of Day 1 & Summary of the Day by the Program Director & The Conference Chair
20.06.2022
17:45 - 17:50 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Sarah FarleyPortfolio Director Smart Mobility & Automotive
we.CONECT Global Leaders GmbH

Perception Tech + Sensor Systems Workshop
AUTO.AI DRIVING DAYS – DEEP DRIVE with our partner Ambarella
20.06.2022
17:50 - 19:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Robert BloomquistVice President of Automotive Business Development
Ambarella
Paul DentelSenior Technical Product Manager
Ambarella

Ambarella’s test drives will feature an SUV outfitted with the following four
technology demonstrations…

 

Forward-Camera AI Demos:

 * One running Ambarella’s AmbaNet neural network
 * Another running Autobrains’ neural network
 * Oculii™ 360-Degree High-Resolution Radar Perception
 * Full Display, 2MP Rearview eMirror




Networking
BITS, BITES AND BUBBLES I The Joint Auto.AI + OSS.5 Dinner
20.06.2022
19:30 - 22:00 (GMT-05:00) Eastern Time (US & Canada)
Dinner Location
 * Do you want to network with your peers?
 * Do you want to digest the day and your key findings?
 * Do you want to sit back & connect with your new acquaintances?

Relax after the first conference day and begin a great evening with drinks, food
and bubbly networking at a special location.


Networking
Registration & Auto.AI Day 2 Community Wall
21.06.2022
07:45 - 08:20 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection

The official registration for Day 2: Grab a coffee, networking with your
community, use your personalized matchmaking app, make next contacts, plan your
meetings and get started. Welcome to Day 2 of the Auto.AI USA!


Presentation
Welcome & Introduction by the Program Director & the Chair
21.06.2022
08:50 - 09:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Sarah FarleyPortfolio Director Smart Mobility & Automotive
we.CONECT Global Leaders GmbH

Presentation
Keynote I Developing reliable Autonomous Driving Behavior capabilities- How cars
can learn from real world driving behavior
21.06.2022
09:00 - 09:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Shweta ShrivastavaSenior Product Leader Autonomous Driving Capabilities
Waymo

The field of autonomous driving behavior poses a major challenge in the
development of AVs. Waymo has been a forerunner in delivering safe and reliable
autonomous driving experiences through its products. I this session we will take
a closer look on how Waymo’s vehicles developed its driving behavior patterns on
the example of pedestrians and cyclists.

 * Waymo’s mission to make it safe and easy for people and things to get where
   they’re going
 * Benefits of focusing on fully automated driving
 * Keys to successful product management
 * How Waymo is learning from real-world applications, including their fleet in
   Phoenix


Perception Tech + Sensor Systems Presentation Sensor Fusion + Computer Vision
Software Engineering
Solution Study I Breaking the Fundamental Robust Perception Gaps for ADAS & AV
21.06.2022
09:30 - 10:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Dave TokicVP Marketing and Strategic Partnerships
Algolux Inc.

ADAS and AV platforms require accurate and robust computer vision (CV) in order
to deliver safe operation. Unfortunately as seen by recent reports from the AAA,
Tesla at CES2022, and direct experiences from OEMs and Tier 1s are
disheartening. There continues to be significant gaps in the practical operating
effectiveness in all conditions of the latest ADAS and AV systems deployed.



Algolux will present multiple proven best-in-industry approaches to

 1. optimize existing vision systems to significantly improve vision performance
 2. rearchitect current methods to massively improve CV robustness in harsh
    conditions, and
 3. deliver unprecedented robust accuracy of dense depth perception.


Ai + Machine Learning + Deep Learning Data Perception Tech + Sensor Systems
Presentation Sensor Fusion + Computer Vision Validation + Simulation + Modeling
Case Study I Lessons Learned from the Mass Deployment of Machine Learning Models
for AV Perception
21.06.2022
10:00 - 10:30 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Dr. Anurag GanguliVice President, Research & Development
Plus

The PlusDrive system is a highly automated driver-in product for heavy duty
trucks that is currently being used across the United States. PlusDrive includes
one of the most, if not the most, complex perception system deployed in a
product for heavy duty trucks consisting of multiple cameras, radars and lidars.
In this talk, we will discuss some of the lessons learned from the mass
deployment of machine learning models for AV perception touching upon runtime
optimization and online learning from data at scale.

 * Approaches for deep model compression and acceleration
 * Effect of structured pruning for autonomous visual perception tasks
 * Transformers for computer vision and the related acceleration approaches


Data Perception Tech + Sensor Systems Presentation Software Engineering
Trajectory Planning + Prediction + Decision-making
Solution Study I When Milliseconds Count
21.06.2022
10:30 - 11:00 (GMT-05:00) Eastern Time (US & Canada)
Plenum
The Impact of Lidar on Perception Improvements
Mrinal SoodDirector of Technical Marketing
AEye

In order for autonomous systems to coexist with human drivers, they need better
reaction times for improved path planning. Existing hardware-centric lidars
offer redundancy, but don’t improve the reaction time perception engineers look
for in order to allow operation at higher speeds.In this session, AEye’s Mrinal
Sood will talk about how adaptive, high performance, software-configurable lidar
allows higher-speed vehicle operation with more graceful braking and throttling
through meaningful long-range data.



Key Takeaways:

 * Hardware-centric lidars today offer redundancy, but have inherent limitations
   for feature improvements
 * The benefits of added reaction time for ADAS systems
 * How adaptive, high performance, software configurable lidar can provide
   meaningful long range data
 * The impact of long-range classification for prediction and path planning and
   how they result in smoother driving at highway speeds


Presentation Workshop
Introduction to World Cafés and short presentation of moderators
21.06.2022
11:00 - 11:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum

Each moderator will showcase in a 3 minutes pitch, his Top 5 challenges needs to
be discussed in his World Cafe Session.


Networking
Breakfast Break & Networking
21.06.2022
11:10 - 11:40 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Get coffee, get snacks, make connections and talk with your peers – and have the
chance to actively participate in product and service demos on the expo floor.


Networking
100 MINUTES – ONE2ONE MEETING SESSIONS (07 | 08)
21.06.2022
11:10 - 11:40 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Through our AI-based matchmaking technology we identify and determine interests,
consultation requirements, consulting expertise and solutions as well as
currently ongoing projects of the participants. This assessment allows for
qualified and time-efficient One2Ones with your colleagues and/or potential
partners. Each session lasts about 10 minutes.

Start of Session 07 & 08 – Search – Find – Match – Network


Software Engineering Validation + Simulation + Modeling Workshop
1 I V+V Café - Challenges to Verification and Validation of L4+ Vehicles
21.06.2022
11:40 - 15:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Irene PeraliSystems Engineering Manager
Cruise Automation
 * How much testing is sufficient that an autonomous vehicle could be considered
   as safe?
 * How to handle validation without a driver?
 * How to test hardware-software integrated performance?
 * How to address software updates in an autonomous vehicle?


Ai + Machine Learning + Deep Learning Workshop
2 I Deep Learning Café - What are the guidelines for developing machine learning
components?
21.06.2022
11:40 - 15:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Ramesh SethuSenior Technical Fellow
General Motors
 * With what challenges the major actors in AD deal to develop ML components?
 * What tools are used to develop ML components?
 *  What are the techniques to verify and validate machine learning?
 *  Is there a need to include the operational domain in ML cycle?


Data Validation + Simulation + Modeling Workshop
3 I CloudFactory Café - How Market Leaders are Adapting to Changing AV Data
Annotation Techniques and Trends
21.06.2022
11:40 - 15:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Paul ChristiansonVP of Strategic Industries
CloudFactory
 * How do you decide where to deploy humans and automation in your existing
   model pipeline?
 * Synthetic data is being increasingly hyped as an option for multiple
   industries. Will synthetic data be usable in the next few years?
 * How do you measure the quality of your model and, if there is an issue, what
   is the strategy for root cause analysis?
 * What data pipeline and model conditions need to be met to reach L4 (or L5)
   automation and removing drivers?
 * Can LiDAR data really be replaced with 2-D cameras and new annotation
   techniques?


Data Validation + Simulation + Modeling Workshop
4 I Data AI Café - What is the Role of Data in AI for Autonomous Driving?
21.06.2022
11:40 - 15:10 (GMT-05:00) Eastern Time (US & Canada)
Virtual Meeting
Oleg GusikhinTechnical Leader
Ford Motor Company
 * How to set up the optimal infrastructure for the data pipeline?
 * How to build the proper architecture in terms of a business viable
   proposition?
 * How to we upload / download unstructured data, especially video data? What
   are the best approaches to label & auto-label video data?
 * What are technical requirements for video data? How do we scale the
   infrastructure? How to address the PI / personal data: How to protect these
   information? How do we anonymize the data? How do we gain value from these
   data? What models do we use? 


Ai + Machine Learning + Deep Learning Connectivity + V2x Data Perception Tech +
Sensor Systems Workshop
5 I Sensing & Perception Café - How can we objectively evaluate alternative
Solutions for Autonomous Vehicles?
21.06.2022
11:40 - 15:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Anthony R. GieseySVP of Automotive & Mobility
Evonomy
 * Which sensors do you use and what are the major obstacles of current
   solutions for autonomous vehicles?
 * From corporate V2X to a completely new sensor suite – What are potential
   alternative solutions?
 * What are the current and future opportunities of machine learning for
   optimizing sensing and perception?
 * How to effectively assess the whole variety of solutions?


Networking
LUNCH BREAK & NETWORKING
21.06.2022
12:40 - 13:40 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Get lunch, get snacks, make connections and talk with your peers – and have the
chance to actively participate in product and service demos on the expo floor.


Networking
100 MINUTES – ONE2ONE MEETING SESSIONS (09 | 10)
21.06.2022
12:40 - 13:40 (GMT-05:00) Eastern Time (US & Canada)
Exhibition Area

Through our AI-based matchmaking technology, we identify and determine
interests, consultation requirements, consulting expertise and solutions as well
as currently ongoing projects of the participants. This assessment allows for
qualified and time-efficient One2Ones with your colleagues and/or potential
partners. Each session lasts about 10 minutes.

Start of Session 09 & 10 – Search – Find – Match – Network


Networking
Afternoon Break & Networking
21.06.2022
15:10 - 15:40 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection

Get coffee, get snacks, make connections and talk with your peers – and have the
chance to actively participate in product and service demos on the expo floor.


Presentation
World Cafe Harvesting Session
21.06.2022
15:40 - 16:10 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection
Irene PeraliSystems Engineering Manager
Cruise Automation
Ramesh SethuSenior Technical Fellow
General Motors
Paul ChristiansonVP of Strategic Industries
CloudFactory
Oleg GusikhinTechnical Leader
Ford Motor Company
Anthony R. GieseySVP of Automotive & Mobility
Evonomy

At the end of the last session (session 5) the moderators will give a short
summary of each world café in front of the audience to showcase all key take
aways.  

All results, pictures and key takeaways of the World Cafes will be photographed
after the event and made available to the we.CONECT Media Center as well as to
our event app “hubs.”


Ai + Machine Learning + Deep Learning Data Perception Tech + Sensor Systems
Presentation
Case Study I Challenges to Machine Learning Empowered Sensor Data Sharing for
Connected and Automated Driving
21.06.2022
16:10 - 16:40 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Hongsheng LuPrincipal Researcher
Toyota North America

This presentation overviews the recent progress on sensor data sharing. It aims
to develop an in-depth understanding of the benefits and challenges of
performing sensor data sharing in the context of connected automated systems.
Highlights will be given to the role of machine learning in building such a
system.

 * Why sensor sharing for automated vehicles?
 * Key challenges in designing a sensor sharing system
 * Machine learning as a building block for sensor sharing


Perception Tech + Sensor Systems Presentation Software Engineering
Case Study I Objectively Assessing Alternative Sensing & Perception Solutions
for Autonomous Vehicles
21.06.2022
16:40 - 17:10 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Anthony R. GieseySVP of Automotive & Mobility
Evonomy

Utilizing insights from Evonomy’s Data Ecosystem which works with over 400
suppliers, the presentation will present the approach to objectively assessing
sensing & perception solutions by explaining the Evonomy Component Selection
Recommendation (CSR) service, specifically covering the Sensing & Perception
area including solutions like Artificial Intelligence SoCs, LiDARs, Radar,
Ultrasonics, Cameras, Software Stacks, etc.

 * How OEMs & Tier-1s set criteria for objectively assessing alternative sensing
   & perception solutions
 * How optimal solutions are identified which meet OEM / Tier-1 criteria,
   including AI SoCs, LiDAR, Radars, etc.
 * How Evonomy’s Component Selection Recommendation (CSR) service supports this
   process


AUTO SCALE START UP LOUNGE
21.06.2022
17:10 - 17:40 (GMT-05:00) Eastern Time (US & Canada)
The Henry Autograph Collection
Emily LeungChief Revenue Officer
Hyperspec AI
Karsten KrispinCEO
rabbitAI

Start-Ups – Innovations – How can I get involved?

You are looking for new technologies in the field of self-driving cars? Your
company is interested in new industry collaborations? You’re searching for the
right start-ups to help you really get things rolling? Look no further: Auto
Scale puts you in touch with start-ups who have what it takes to put your
business on the right track. Listen to 5 carefully selected start-ups who
present their golden dust in a 3 minutes Pecha Kucha Pitch! We‘re bringing
creative talents and experienced managers from the automotive industry together.
Network, brainstorm and join the catalysts for self-driving cars start-ups –
that‘s Auto Scale at Tech.AD Europe.Bringing creative talents together with
experienced managers from the automotive industry, network & brainstorm, join
the catalyst for self-driving cars start-ups – that’s Auto Scale at Auto Scale
at Tech.AD Europe!


Presentation
End of Day 2 & Summary of the Day by the Program Director & The Conference Chair
21.06.2022
17:40 - 17:50 (GMT-05:00) Eastern Time (US & Canada)
Plenum
Sarah FarleyPortfolio Director Smart Mobility & Automotive
we.CONECT Global Leaders GmbH

INFORMATION

There is no session that matches your filter criteria

Book your ticket!



BOOK NOW BOOK AS TEAM
 * About we.CONECT
 * we.People
 * Contact



BOOK NOW BOOK AS TEAM
 * Media & Webcasts
 * Digital Managed Events
 * Live Events
 * Virtual Events


BOOK NOW BOOK AS TEAM
© we.CONECT GLOBAL LEADERS - Tomorrow's Business Company
 * Imprint
 * Terms & Conditions
 * Privacy by design

© we.CONECT GLOBAL LEADERS - Tomorrow's Business Company

[wpforms title="true" description="true" id="
Privacy Preference

We use cookies on our website. Some of them are essential, while others help us
to improve this website and your experience.

 * Essential
 * Marketing
 * External Media

Accept all

Save

Individual Privacy Preferences

Cookie Details Privacy Policy Imprint

Privacy Preference

Here you will find an overview of all cookies used. You can give your consent to
whole categories or display further information and select certain cookies.

Accept all Save

Back

Essential (9)


Essential cookies enable basic functions and are necessary for the proper
function of the website.

Show Cookie Information Hide Cookie Information

Name Borlabs Cookie Provider Owner of this website Purpose Saves the visitors
preferences selected in the Cookie Box of Borlabs Cookie. Cookie Name
borlabs-cookie Cookie Expiry 1 Year

Name Polylang Provider Owner of the website Purpose Saves information about the
selected language setting. Privacy Policy https://polylang.pro/privacy-policy/
Cookie Name pll_language Cookie Expiry 1 Jahr

Name Wordpress (Logged in) Provider Owner of this website Purpose Wordpress
cookie which is used to keep users logged in. Privacy Policy
https://wordpress.org/about/privacy/ Cookie Name wordpress_logged_in_*

Name Wordpress (Security) Provider Owner of this website Purpose Wordpress
cookie, which is saved to protect against hackers. Privacy Policy
https://wordpress.org/about/privacy/ Cookie Name wordpress_sec_*

Name Wordpress (Settings) Provider Owner of this website Purpose Wordpress
cookie to save the properties of the user logged on to the backend. Privacy
Policy https://wordpress.org/about/privacy/ Cookie Name wp-settings-*

Name Wordpress (Settings Time) Provider Owner of this website Purpose The
Wordpress cookie is used to adapt your view of the admin interface and possibly
also the main interface. Privacy Policy https://wordpress.org/about/privacy/
Cookie Name wp-settings-time-*

Name WPForms Provider Owner of this website Purpose The cookie is set by the
WPForms plugin. It saves the UserID of the user logged on to the backend.
Privacy Policy https://wpforms.com/privacy-policy/ Cookie Name _wpfuuid

Name Wordpress (Test Cookie) Provider Owner of this website Purpose Is set by
Wordpress to test whether a cookies can be saved. Privacy Policy
https://wordpress.org/about/privacy/ Cookie Name wordpress_test_cookie

Name Register Provider Owner of this website Purpose After a successful
registration for participation in an event or for downloading documents this
cookie is saved and is intended to prevent the user from logging in again.
Privacy Policy https://privacy.we-conect.com/privacy-policy/ Cookie Name
qq33d1ad3aggtoe

Marketing (3)


Marketing cookies are used by third-party advertisers or publishers to display
personalized ads. They do this by tracking visitors across websites.

Show Cookie Information Hide Cookie Information

Accept Name Google Analytics Provider Google LLC Purpose Cookie by Google used
for website analytics. Generates statistical data on how the visitor uses the
website. Privacy Policy https://policies.google.com/privacy?hl=en Cookie Name
_ga,_gat,_gid Cookie Expiry 2 Years

Accept Name Zlc - Cookies Provider Owner of the website Purpose The cookie
enables you to continue a chat while you are browsing different pages on our
website or when you return to the website later. Privacy Policy
https://www.zendesk.com/company/customers-partners/privacy-policy/ Cookie Name
__zlcmid, __zlcprivacy

Accept Name Mail Tracking Provider Owner of the website Purpose Stores
information about whether a user clicked the link to the website via an email
sent by we-conect. Privacy Policy https://privacy.we-conect.com/privacy-policy/
Cookie Name trakingToken

External Media (1)


Content from video platforms and social media platforms is blocked by default.
If External Media cookies are accepted, access to those contents no longer
requires manual consent.

Show Cookie Information Hide Cookie Information

Accept Name YouTube Provider YouTube Purpose Used to unblock YouTube content.
Privacy Policy https://policies.google.com/privacy?hl=en&gl=en Host(s)
google.com Cookie Name NID Cookie Expiry 6 Monate

powered by Borlabs Cookie

Privacy Policy Imprint