Autonomous driving represents a significant advancement in the transportation industry, enhancing vehicle intelligence, optimizing traffic management, and improving user experiences. Central to these innovations is deep learning, which enables systems to handle complex data and make informed decisions. Our survey explores critical applications of deep learning in autonomous driving, such as perception and detection, localization and mapping, and decision-making and control. We investigate specialized deep learning techniques, including convolutional neural networks, recurrent neural networks, self-attention transformers, and their variants, among others. These methods are applied within various learning paradigms—supervised, unsupervised, and reinforcement learning—to suit the specific needs of autonomous driving. Our analysis evaluates the effectiveness, benefits, and limitations of these technologies, focusing on their integration with other intelligent algorithms to enhance system performance. Furthermore, we examine the architectures of autonomous systems, analyzing how knowledge and information are organized from modular, pipeline-based frameworks to comprehensive end-to-end models. By presenting an exhaustive overview of the progressing domain of autonomous driving and bridging various research areas, our survey aims to synthesize diverse research threads into a unified narrative. This effort not only aims to enhance our understanding but also pushes the boundaries of what is achievable in this interdisciplinary field.

Vehicular crowd sensing systems are designed to achieve large spatio-temporal sensing coverage with low-cost in deployment and maintenance. For example, taxi platforms can be utilized for sensing city-wide air quality. However, the goals of vehicle agents are often inconsistent with the goal of the crowdsourcer. This inconsistency decreases the sensing coverage quality and thus impairs the quality of the collected information. A simple approach to reduce the inconsistency is to greedily incentivize the vehicle agents to different regions. However, incentivization brings challenges, including the heterogeneity of desired target distributions, limited budget to incentivize more vehicle agents, and the high computational complexity of optimizing incentivizing strategies. To this end, we present a vehicular crowd sensing system to efficiently incentivize the vehicle agents to match the sensing distribution of the sampled data to the desired target distribution with a limited budget. To utilize the budget efficiently, we design a customized incentive by combining monetary incentives and potential task(ride) requests at the destination. Meanwhile, an efficient optimization algorithm, iLOCuS, is presented to plan the incentivizing policy for vehicle agents to decompose the sensing distribution into two distinct levels. Our experimental results based on real-world data show that our system reduced 26.99% dissimilarity.

Fine-grained city-scale outdoor air pollution maps provide important environmental information for both city managers and residents. Installing portable sensors on vehicles (e.g., taxis, Ubers) provides a low-cost, easy-maintenance, and high-coverage approach to collecting data for air pollution estimation. However, as non-dedicated platforms, vehicles like taxis usually prefer gathering at busy areas of a city where it is more likely to pick up riders. This leaves many parts of the city unsensed or less-sensed. In addition, due to the natural changes in a city and the movements of the vehicles, the sensed and unsensed areas change over time. Consequently, challenges of air pollution estimation with data collected by non-dedicated mobile platforms are twofold: i. data coverage is sparse; ii. data coverage changes over time. Therefore, the major research question is: how can we derive accurate and robust fine-grained field (e.g., air pollution) estimation given dynamic and sparse data collected from uncontrollable mobile sensing platforms? This paper presents adaptive HMSS , an adaptive h ybrid m odel-enabled s ensing s ystem for fine-grained air pollution estimation with dynamic and sparse data collected from uncontrollable mobile sensing platforms, which is achieved by combining the advantages of a physics guided model and a data driven model . To address the challenge of sparse coverage, the physical understanding of the spatiotemporal correlation for air pollution distribution in the physics guided model is utilized to infer values at unsensed sparse areas. Meanwhile, the data driven model is adopted to estimate the air pollution influential factors (e.g., buildings) not included in the physics guided model . To address the challenge of time-varying coverage, an adaptive model combination algorithm is designed to enable the system bias to either of the two models according to the amount of data collection and uncertainty of the model. To evaluate the system performance, we deployed 47 air pollution sensing devices on taxis and fixed locations in 2 cities for both controlled and uncontrolled experiments for over two weeks. The results show that with a resolution of $500 \;\mathrm m$ by $500\;\mathrm m$ by $1\;\mathrm {hour}$ , our system achieves up to $3.2\times$ error reduction when compared to the baseline approaches.

Rapid post-earthquake reconnaissance is important for emergency responses and rehabilitation by providing accurate and timely information about secondary hazards and impacts, including landslide, liquefaction, and building damage. Despite the extensive collection of geospatial data and satellite images, existing physics-based and data-driven methods suffer from low estimation performance due to the complex and event-specific causal dependencies underlying the cascading processes of earthquake-triggered hazards and impacts. Herein, we present a rapid seismic multi-hazard and impact estimation system that leverages advanced statistical causal inference and remote sensing techniques. The unique feature of this system is that it provides accurate and high-resolution estimations on a regional scale by jointly inferring multiple hazards and building damage from satellite images through modeling their causal dependencies. We evaluate our system on multiple seismic events from diverse countries around the globe. Our results corroborate that incorporating causal dependencies significantly improves large-scale estimation accuracy for multiple hazards and impacts compared to existing systems. The results also reveal quantitative causal mechanisms among earthquake-triggered multi-hazard and impact for multiple seismic events. Our system establishes a new way to extract and utilize the complex interactions of multiple hazards and impacts for effective disaster responses and advancing understanding of seismic geological processes.

Multi-UAV systems have shown immense potential in handling complex tasks in large-scale, dynamic, and cold-start (i.e., limited prior knowledge) scenarios, such as wildfire suppression. Due to the dynamic and stochastic environmental conditions, the scheduling for sensing tasks (i.e., fire monitoring) and operation tasks (i.e., fire suppression) should be executed concurrently to enable real-time information collection and timely intervention of the environment. However, the planning inclinations of sensing and operation tasks are typically inconsistent and evolve over time, complicating the task of identifying the optimal strategy for each UAV. To solve this problem, this paper proposes SOScheduler, a collaborative multi-UAV scheduling framework for integrated sensing and operation in large-scale and dynamic wildfire environments. We introduce a spatio-temporal confidence-aware assessment model to dynamically and directly pinpoint locations that can optimally enhance the understanding of environmental dynamics and operational effectiveness, as well as a priority graph-instructed scalable scheduler to coordinate multi-UAV in an efficient manner. Experiments on real multi-UAV testbeds and large-scale physical feature-based simulations show that our SOScheduler reduces the fire expansion ratio by 59% and enhances the fire coverage ratio by 190% compared to state-of-the-art (SOTA) solutions.