Skip to main content
Welcome to the cutting edge of safety science—Learn more about our rebrand.
  • Training

Machine Learning Safety Training for Autonomous Vehicles

Create safer AI-based mobility

Smart Car sensing people walking

Course overview

Artificial intelligence (AI) plays a significant role in the automotive industry as it competes to develop and market vehicles that can perform tasks that previously required an experienced driver. Nowadays, carmakers and suppliers must face the difficulties of developing all the operational and tactical functions required to operate a vehicle in on-road traffic that performs driving maneuvers better than an experienced driver.

This machine learning (ML) for automotive safety assurance training provides guidance for the safe development of ML components into the state-of-the-art safety frameworks used for automotive, including functional safety (FuSa) and safety of the intended functionality (SOTIF). This training also includes the reference information of novel autonomy safety standards such as:

  • UL 4600, the Standard for Evaluation of Autonomous Products.
  • ISO/TR 4804:2020, Safety and cybersecurity for automated driving systems - Design, verification and validation.
  • ISO/IEC AWI TR 5469, Artificial intelligence - Functional safety and AI systems.
  • ISO/AWI PAS 8800, Road Vehicles - Safety and artificial intelligence.
  • ISO/IEC TR 24029-1:2021, Artificial Intelligence (AI) - Assessment of the robustness of neural networks - Part 1: Overview.
  • ISO/IEC CD 24029-2, Artificial intelligence (AI) - Assessment of the robustness of neural networks - Part 2: Methodology for the use of formal methods.

What is covered in this training?

Day 1

  • Intro
  • Automotive safety and AI standards.
  • Deployment of AI into systems.
  • Factors to consider for AI safety.
  • Running example.
  • Specification and design (SOTIF-specific).
  • AI safety culture.
  • Concept phase.
  • System development phase.

Day 2

  • Hardware development phase.
  • Software development phase.
  • Verification and validation (V&V) strategy.
  • Software verification phase.
  • Hardware verification phase.
  • System verification phase.
  • Vehicle verification and validation phase.
  • Assessment
  • Operations, decommissioning, service and maintenance.

Why should we take this training?

Upon successful completion of this workshop, you will be able to:

  • Apply ML concepts within FuSa and SOTIF frameworks.
  • Develop safer, more robust ML algorithms to solve autonomy safety problems.
  • Obtain sufficient guidance to define, specify, develop, evaluate, deploy and monitor ML algorithms in the context of autonomy safety.
  • Provide ML safety analysis that supports completeness and correctness for safer autonomous vehicles to an acceptable society risk level.
  • Attempt the UL Certified Autonomy Safety Professional (UL-CASP) in Machine Learning exam.

Who is this training for?

  • Engineers working with advanced driver assistance systems (ADAS) and autonomous vehicles (AV), such as automotive engineers, safety engineers, data scientists, project leaders, testing personnel.

Prerequisites

  • UL Certified Artificial Intelligence Professional –Foundations training or hands-on experience developing ML models.
  • Previous knowledge of SOTIF (ISO 21448) and Automotive functional safety (ISO 26262).

Why choose UL Solutions?

From materials testing to supply chain management, new energy options to security and interoperability solutions, leverage our expertise and insights to navigate the global regulatory landscape and bring your products to market.

UL Solutions’ global network of technical experts and state-of-the-art facilities helps manufacturers gain  the knowledge they need to achieve their compliance credentials that are essential to compete in a more complex global supply chain.

Knowledge you can trust – Our experienced staff can support you from the initial design stage of product development through testing and production. Our experts can assist you in understanding the certification requirements for your specific markets.

Speed and efficiency – Our cost-effective systems and state-of-the-art facilities can help accelerate your time to market.

Single-source provider – UL Solutions can help you meet  all your compliance needs and, by bundling safety, performance and interoperability services, can also help you save valuable time and money.

Global reach and access – Our global network of expert engineers can help you understand the various national and global requirements for your specific market application.

Related standards

UL 4600 Standard for Evaluation of Autonomous Products
ISO/IEC TR 24029-1:2021 – Artificial Intelligence
ISO/AWI PAS 8800 Road Vehicles — Safety and artificial intelligence 

Download our resource

UL

UL Certified Autonomy Safety Professional Training In Machine Learning

1.74 MB