English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Autonomous driving systems (ADSs) are capable of sensing the environment andmaking driving decisions autonomously. These systems are safety-critical, andtesting them is one of the important approaches to ensure their safety.However, due to the inherent complexity of ADSs and the high dimensionality oftheir operating environment, the number of possible test scenarios for ADSs isinfinite. Besides, the operating environment of ADSs is dynamic, continuouslyevolving, and full of uncertainties, which requires a testing approach adaptiveto the environment. In addition, existing ADS testing techniques have limitedeffectiveness in ensuring the realism of test scenarios, especially the realismof weather conditions and their changes over time. Recently, reinforcementlearning (RL) has demonstrated great potential in addressing challengingproblems, especially those requiring constant adaptations to dynamicenvironments. To this end, we present DeepQTest, a novel ADS testing approachthat uses RL to learn environment configurations with a high chance ofrevealing abnormal ADS behaviors. Specifically, DeepQTest employs DeepQ-Learning and adopts three safety and comfort measures to construct the rewardfunctions. To ensure the realism of generated scenarios, DeepQTest defines aset of realistic constraints and introduces real-world weather conditions intothe simulated environment. We employed three comparison baselines, i.e.,random, greedy, and a state-of-the-art RL-based approach DeepCOllision, forevaluating DeepQTest on an industrial-scale ADS. Evaluation results show thatDeepQTest demonstrated significantly better effectiveness in terms ofgenerating scenarios leading to collisions and ensuring scenario realismcompared with the baselines. In addition, among the three reward functionsimplemented in DeepQTest, Time-To-Collision is recommended as the best designaccording to our study.

DeepQTest: Testing Autonomous Driving Systems with Reinforcement Learning and Real-world Weather Data