Abstract The generation of corner cases has become increasingly crucial for efficiently testing autonomous vehicles prior to road deployment. However, existing methods struggle to accommodate diverse testing requirements and often lack the ability to generalize to unseen situations, thereby reducing the convenience and usability of the generated scenarios. A method that facilitates easily controllable scenario generation for efficient autonomous vehicles (AV) testing with realistic and challenging situations is greatly needed. To address this, OmniTester is proposed as a multimodal Large Language Model (LLM) based framework that fully leverages the extensive world knowledge and reasoning capabilities of LLMs. OmniTester is designed to generate realistic and diverse scenarios within a simulation environment, offering a robust solution for testing and evaluating AVs. In addition to prompt engineering, OmniTester employs tools from Simulation of Urban Mobility to simplify the complexity of codes generated by LLMs. It further incorporates Retrieval-Augmented Generation and a self-improvement mechanism to enhance the LLM's understanding of scenarios, thereby increasing its ability to produce more realistic scenes. Experiment results demonstrated the controllability and realism of the proposed approaches in generating three types of challenging and complex scenarios. Additionally, OmniTester effectively reconstructs novel scenarios described in crash reports, driven by the generalization capability of LLMs.

Abstract The primary goal of artificial intelligence (AI) is to learn, mimic, and ultimately surpass human intelligence. Connectionist AI draws inspiration from human neuroscience and biological mechanisms and has made significant progress in many fields like autonomous driving (AD). However, the performance of AD systems in open, complex, and dynamic environments is still limited. Integrating robust and high-level human intelligence into AI remains a critical trend for its ongoing development and evolution. Inspired by this, this paper reviews Human Intelligence Augmented Artificial Intelligence (HIA-AI) from an AD perspective and offers a taxonomy of related methods, including learning from human demonstrations, tuning from human feedback, integrating from human mechanisms, and abstracting from human knowledge. It discusses current applications and innovations of related methods. Furthermore, this paper examines the characteristics, advantages, and disadvantages of various HIA-AI methods and points out potential future research directions.

As a promising clean energy conversion device, solid oxide fuel cells (SOFCs) could efficiently utilize multiple fuels and have been applied in stationary power generation and other fields. However, the high operating temperature limits the use of SOFCs in the transportation sector. Consequently, reducing the operating temperature is a key focus in SOFCs development. In recent years, metal-supported SOFCs (MS-SOFCs) have received significant attention because of intermediate operating temperatures. MS-SOFCs offer advantages such as simple and reliable sealing, rapid start-up, and high power density, making MS-SOFCs a promising option for transportation applications. However, for commercial applications, the methods to enhance performance, improve durability and mitigate degradation need to be further studied. This paper comprehensively reviews the state of the art approaches to performance improvement, durability under various conditions and different fuel reforming methods. Finally, it highlights the prospects and challenges for future advancements in this field.