Main Day Two | Thursday, April 28, 2022

Recharge your batteries with breakfast and prepare for the day ahead!

Failure of the car is not just a safety problem; it is a problem of brand image. The automotive industry cannot afford any errors.

Kick off day two with a look to the AV bubble. The AV industry is innovating at an unprecedented speed, but Level-4 AVs remain untamed.

Hear the necessity behind SOTIF in enforcing our accepted level of safety, testing and certainty in AVs, allowing for progressive testing and analysis in our road to level four vehicles in human scenarios.

• Should vs Must: the need for urgent action in implementing SOTIF and safety infrastructure in ADs and ADAS product design
• Appreciate the risk of putting products to market too early and causing unnecessary incidents
• The AV industry’s fate at the mercy of funding and the court of public opinion, the industry cannot suffer a major autonomous incident, or all hope is lost for all

Dr. Håkan Sivencrona

Chief Safety Officer
Zenseact

See Bio

Ali Nouri

Senior System Safety Engineer
Volvo Cars

See Bio

When something is unknown, how are you going to accept the risks? Because risk is a measure of uncertainty.

Along with accepting residual risk is the concept of arguing for sufficient safety. Testing, analysis and evaluation will provide data and simulated scenarios to be considered in product design, but ultimately it comes down to the industry to debate what is sufficient.

Join fellow industry experts and debate the question: How good do we really have to be?

• At what point is safety considered good enough?
• How do you define good enough?
• What can we do to create a benchmark for sufficient safety?
• What technological advancements are needed in tightening the threshold for sufficient safety?

Dr. Håkan Sivencrona

Chief Safety Officer
Zenseact

See Bio

Dr Christian Schwarzl

Head of Dependable Systems Group - Electrics, Electronics and Software
Virtual Vehicle Research Center

See Bio

Simone Fabris

Senior Director and General Manager - Compute Platform Industrialization at Mobileye
Mobileye

See Bio

Antonio Priore

Director of Functional Safety and Cybersecurity
ARM

See Bio

Ali Nouri

Senior System Safety Engineer
Volvo Cars

See Bio

Refuel with a coffee, and catchup with fellow attendees!

To what extent are we keeping the separation between functional safety and safety of the intended functionality? Does that distinction continue to make sense?

Calls have come from within the 26262 community to expand the standard to include AI and ML, so when do we address the elephant in the vehicle, is it worth merging 21448 and 26262? Doing so would streamline regulatory understanding and condense discipline knowledge. But FuSa is not subject to the infinite errors that are seen with SOTIF. 

Join two industry experts as they discuss and answer burning questions on the nature of the relationship between SOTIF and FuSa, ISO 21448 and IS0 26262. Including: 

• Harmonizing SOTIF and FuSa, what would it look like?
• How do we ease the workflow in identifying and practicing FuSa and SOTIF simultaneously?
• Debating overcoming the different levels of scenarios seen, structured vs infinite

Nicolas Becker

Project Leader ISO 21448 'SOTIF' Standard
Stellantis

See Bio

When we talk about SOTIF, we're talking about technology limitations. The part that is the reason why ADs cars are not already deployed is the lack of safe artificial intelligence algorithms. Safety is not trivial.

Ahead of our joint session with Application of ISO 26262 on making AI fully safe, this session takes a look at the role AI will and can play in regard to SOTIF. Ultimately SOTIF is limited by the advance of the partnered technology in testing for unknown unsafe scenarios and simulating hazards, subsequently leaving a great deal of residual risk. AI presents the opportunity to increase our understanding and knowledge of those unknown hazards and better build our components, and eventually achieve level 4 ADs. 

Join this presentation to understand the role AI has to play in SOTIF

• How AI can help build more diversified data sets for our safety algorithms
• What effective AI can do to aid in validation and verification of Area 2 and Area 3 evaluations in SOTIF analysis
• The plans for the AI and SOTIF workgroup to achieve standardization and harmonization between the two components

But also, how to make your AI SOTIF compliant

Suraj Chandwani

Functional Safety Expert
ION Energy

See Bio

 Hear from an industry expert and discuss:

• With little to no guidance on safety standards for AI and ML, what requirements can the industry benchmark against and work towards?
• Exchanging recommendations and ideas on the ultimate safety requirements for AI, ML and deep learning solutions
• Can AI be fully trusted by the driver? What is the role of the driver when applying functional safety to AI assisted systems
• Is AI the best solution? Discuss the long term role of AI in the future of autonomous vehicles

Kathrin Kind-Trueller

Senior Research Scientist in AI for Vehicle Control
Volkswagen

See Bio

Touch base with your team members, make new connections or both! Get chatting over lunch

In this session, join a safety expert from a market leading Tier1 who will present best practices for complying with functional safety at the software level.

• Achieving full functional safety for integrated software architectures
• Impact of increased software complexity on qualification processes
• Software safety analysis – All you can eat?
• Specific challenges of ML systems

Dr. Wolfgang Freese

Director Functional Safety
Aptiv

See Bio

• Learn about the challenges of integrating software safety components
• Find out best practices for performing system level tests to meet safety requirements
• Discover how integration testing can ease software integration by exposing faults
• Gain an insight into how integration testing can optimise reliability

Paolo Bernardi

Professor
Politecnico di Torino

See Bio

Join this panel to discuss how the industry is addressing new compliance challenges, with the end goal of establishing best practices for ISO 26262 compliant complex software

• Expert recommendations of how to handle existing software
• Best practices for applying existing methodologies to new complex software
• How the industry is adjusting to convergence right across the application domain
• Developments in the compliance of embedded software systems

Paolo Bernardi

Professor
Politecnico di Torino

See Bio

Dr. Wolfgang Freese

Director Functional Safety
Aptiv

See Bio

Dave Higham

VP Functional Safety
Codasip

See Bio

Riccardo Vincelli

Senior Manager, Functional Safety Competence Center
Renesas

See Bio

Join this technical presentation to gain clarity on how ASIL decomposition can be an asset when applied correctly, and discuss:

• How ASIL decomposition allows to relax some requirements in case of redundancy
• How to understand ISO 26262-9 5.4.7
• Expert recommendations for optimizing use of ASIL decomposition
• Optimizing safety architecture through ASIL decomposition

Dr. Pierre Metz

Global Lead Functional Safety
Brose Group

See Bio

Join this panel discussion to see standard experts come to together with members of the industry to discuss the business case for standardisation, and what needs to be done in the real sense to achieve with a streamlined framework of functional safety requirements.

• Find out how standardisation of safety requirements can increase the scope of improvement and innovation in functional safety
• Explore the benefits of being on the same framework when new players enter the industry
• Learn how streamlining functional safety requirements can improve development flow from the hardware to the system level

Darren Galpin

Principal Digital Verification Engineer
Dialog Semiconductors, a Renesas Company

See Bio

Dr. Pierre Metz

Global Lead Functional Safety
Brose Group

See Bio

Dr Simon Hart

Chief Technology Innovation Officer
YASA

See Bio

Riccardo Vincelli

Senior Manager, Functional Safety Competence Center
Renesas

See Bio