Uncertainty-Aware Measurement of Scenario Suite Representativeness for Autonomous Systems

TL;DR

This paper introduces a probabilistic, uncertainty-aware approach to measure how well scenario datasets for autonomous systems reflect real operational conditions, addressing data limitations and prior uncertainties.

Contribution

It proposes a novel imprecise Bayesian method to quantify scenario suite representativeness with interval estimates, accounting for limited data and prior uncertainty.

Findings

Interval-valued representativeness estimates demonstrate uncertainty handling.

Application to autonomous vehicle scenarios across categories shows practical effectiveness.

Method captures local and global representativeness under data constraints.

Abstract

Assuring the trustworthiness and safety of AI systems, e.g., autonomous vehicles (AV), depends critically on the data-related safety properties, e.g., representativeness, completeness, etc., of the datasets used for their training and testing. Among these properties, this paper focuses on representativeness-the extent to which the scenario-based data used for training and testing, reflect the operational conditions that the system is designed to operate safely in, i.e., Operational Design Domain (ODD) or expected to encounter, i.e., Target Operational Domain (TOD). We propose a probabilistic method that quantifies representativeness by comparing the statistical distribution of features encoded by the scenario suites with the corresponding distribution of features representing the TOD, acknowledging that the true TOD distribution is unknown, as it can only be inferred from limited data. We apply an imprecise Bayesian method to handle limited data and uncertain priors. The imprecise Bayesian formulation produces interval-valued, uncertainty-aware estimates of representativeness, rather than a single value. We present a numerical example comparing the distributions of the scenario suite and the inferred TOD across operational categories-weather, road type, time of day, etc., under dependencies and prior uncertainty. We estimate representativeness locally (between categories) and globally as an interval.