Waymo's Safety Methodologies and Safety Readiness Determinations

10/30/2020
by   Nick Webb, et al.
0

Waymo's safety methodologies, which draw on well established engineering processes and address new safety challenges specific to Automated Vehicle technology, provide a firm foundation for safe deployment of Waymo's Level 4 ADS, which Waymo also refers to as the Waymo Driver. Waymo's determination of its readiness to deploy its AVs safely in different settings rests on that firm foundation and on a thorough analysis of risks specific to a particular Operational Design Domain. Waymo's process for making these readiness determinations entails an ordered examination of the relevant outputs from all of its safety methodologies combined with careful safety and engineering judgment focused on the specific facts relevant for a particular determination. Waymo will approve when it determines the ADS is ready for the new conditions without creating any unreasonable risks to safety. This paper explains Waymo's methodologies as applied to the three layers of its technology: hardware, ADS behavior, and operations, and also explains Waymo's safety governance. Waymo will continue to apply and adapt those methodologies, and to learn from the important contributions of others in the AV industry, as Waymo continues to build an ever safer and more able ADS.

READ FULL TEXT
research
06/02/2023

Building a Credible Case for Safety: Waymo's Approach for the Determination of Absence of Unreasonable Risk

This paper presents an overview of Waymo's approach to building a reliab...
research
11/08/2022

System Safety Engineering for Social and Ethical ML Risks: A Case Study

Governments, industry, and academia have undertaken efforts to identify ...
research
06/23/2023

Interpreting Safety Outcomes: Waymo's Performance Evaluation in the Context of a Broader Determination of Safety Readiness

This paper frames recent publications from Waymo within the broader cont...
research
11/15/2021

A Finite-Sampling, Operational Domain Specific, and Provably Unbiased Connected and Automated Vehicle Safety Metric

A connected and automated vehicle safety metric determines the performan...
research
07/31/2020

SafetyLens: Visual Data Analysis of Functional Safety of Vehicles

Modern automobiles have evolved from just being mechanical machines to h...
research
01/14/2022

A causal model of safety assurance for machine learning

This paper proposes a framework based on a causal model of safety upon w...
research
08/03/2022

Fast or Accurate? Governing Conflicting Goals in Highly Autonomous Vehicles

The tremendous excitement around the deployment of autonomous vehicles (...

Please sign up or login with your details

Forgot password? Click here to reset