LAMBDA: Covering the Multimodal Critical Scenarios for Automated Driving Systems by Search Space Quantization

TL;DR

The paper tackles the challenge of comprehensively evaluating automated driving systems when critical scenarios are multimodal and sparsely distributed in a high-dimensional logical space. It formalizes Black-Box Coverage (BBC) as identifying all points with $f(oldsymbol{x})>oldsymbol{\delta}$ under a limited budget, using Confusion Matrix-based $F_2$ as a coverage metric. To solve BBC, it introduces LAMBDA, a density-aware, latent-action Monte-Carlo beam search that partitions the space, uses beam-like parallel exploration, and adapts exploration via a density-based UCB, with integration of BO and TuRBO as local samplers. Across low- and high-dimensional synthetic benchmarks and a practical ADS scenario, LAMBDA demonstrates superior coverage speed, stability, and near-complete modality coverage, offering a compelling tool for virtual safety evaluation. The work advances ADS testing by shifting focus from single-optimum falsification to broad, reliable scenario coverage, enabling more robust safety assessments and potential deployment readiness.

Abstract

Scenario-based virtual testing is one of the most significant methods to test and evaluate the safety of automated driving systems (ADSs). However, it is impractical to enumerate all concrete scenarios in a logical scenario space and test them exhaustively. Recently, Black-Box Optimization (BBO) was introduced to accelerate the scenario-based test of ADSs by utilizing the historical test information to generate new test cases. However, a single optimum found by the BBO algorithm is insufficient for the purpose of a comprehensive safety evaluation of ADSs in a logical scenario. In fact, all the subspaces representing danger in the logical scenario space, rather than only the most critical concrete scenario, play a more significant role for the safety evaluation. Covering as many of the critical concrete scenarios in a logical scenario space through a limited number of tests is defined as the Black-Box Coverage (BBC) problem in this paper. We formalized this problem in a sample-based search paradigm and constructed a coverage criterion with Confusion Matrix Analysis. Furthermore, we propose LAMBDA (Latent-Action Monte-Carlo Beam Search with Density Adaption) to solve BBC problems. LAMBDA can quickly focus on critical subspaces by recursively partitioning the logical scenario space into accepted and rejected parts. Compared with its predecessor LaMCTS, LAMBDA introduces sampling density to overcome the sampling bias from optimization and Beam Search to obtain more parallelizability. Experimental results show that LAMBDA achieves state-of-the-art performance among all baselines and can reach at most 33 and 6000 times faster than Random Search to get 95% coverage of the critical areas in 2- and 5-dimensional synthetic functions, respectively. Experiments also demonstrate that LAMBDA has a promising future in the safety evaluation of ADSs in virtual tests.

Figures (20)

• Figure 1: An example to show the multimodal nature of the distribution of critical scenarios.
• Figure 2: The confusion matrix of BBC problem.
• Figure 3: The evaluation process of BBC problem.
• Figure 4: The workflow of LaMCTS.
• Figure 5: An intuitive example of sampling bias.