Advances in Perception, Planning and Control for Automated Vehicles

Dear Colleagues,

Automated vehicles are progressing from controlled demonstrations toward operation in open, complex, and heterogeneous road environments. Reliable and scalable automation depends on coordinated advances across sensing, localization, perception, prediction, decision-making, planning, vehicle dynamics, and control. These advances may arise from a broad methodological spectrum, including signal processing, geometric and physics-based modelling, Bayesian estimation, optimization, operations research, game theory, and classical control. Meanwhile, emerging artificial intelligence techniques (foundation models, vision–language and vision–language–action models, world and world action models, deep reinforcement learning), multimodal sensing, connected infrastructure, and data-driven control are creating new opportunities to provide important additional capabilities for vehicles to understand complex scenes, anticipate plausible futures, interact with human-driven vehicles and vulnerable road users, and act safely under uncertainty.

Major deployment challenges remain, including sensing degradation and component faults, adverse weather and lighting, long-tail and out-of-distribution scenarios, multimodal uncertainty, complex interactions with human-driven vehicles and vulnerable road users, limited transfer across operational domains, computational constraints, weak interpretability or safety guarantees, and gaps between simulation and real-world performance. Addressing these challenges requires rigorous system integration and validation while accounting for vehicle dynamics, human behaviour, connectivity, infrastructure support, system health, comfort, energy use, and real-time implementation.

This Special Issue aims to present and disseminate recent theoretical, methodological, and application advances in perception, prediction, planning, and control for automated vehicles. We welcome model-based, rule-based, probabilistic, optimization-based, control-theoretic, data-driven, learning-based, simulation-based and hybrid approaches, together with modular, integrated, and end-to-end system studies. State-of-the-art artificial intelligence (including foundation models, world models, vision–language–action models, and reinforcement learning), is particularly welcome when it delivers demonstrable gains in robustness, safety, explainability, efficiency, or deployment readiness; however, the use of artificial intelligence is not a prerequisite.

Topics of interest for publication include, but are not limited to, the following:

• Vehicle sensing, sensor calibration, signal processing, and multimodal or cooperative sensor fusion using cameras, LiDAR, radar, ultrasonic sensors, GNSS/IMU, maps, and V2X information;
• Localization, mapping, simultaneous localization and mapping, map matching, trustworthy state estimation, and cooperative positioning;
• Geometric, physics-informed, probabilistic, optimization-based, and learning-based detection, segmentation, tracking, lane and road understanding, and vulnerable-road-user perception;
• Scene representations and understanding, including occupancy grids, bird's-eye-view representations, 3D/4D reconstruction, semantic maps, and spatiotemporal environment models;
• Behaviour modelling, intent recognition, trajectory prediction, and interaction modelling using rule-based, game-theoretic, probabilistic, data-driven, or hybrid methods;
• Decision-making and motion planning based on graph search, sampling, optimization, optimal control, model predictive control, reachability analysis, control barrier functions, or learning;
• Vehicle dynamics and control, including robust, adaptive, nonlinear, predictive, optimal, shared, and fault-tolerant control;
• Integrated and closed-loop co-design of perception, prediction, planning, and control, including modular, hybrid, and end-to-end architectures;
• Foundation models, multimodal large language models, vision-language models, vision–language–action models, world and world action models, and reinforcement learning for automated driving;
• Self-supervised, weakly supervised, semi-supervised, continual, transfer, federated, and domain-generalized learning;
• Uncertainty quantification, out-of-distribution and anomaly detection, fault diagnosis, condition monitoring, prognostics and health management, and graceful degradation;
• Connected and cooperative automated driving, mixed-autonomy coordination, platooning, vehicle-infrastructure collaboration, and distributed control;
• Advanced simulation methodologies and platforms for automated driving, including multi-agent traffic simulation, sensor and vehicle dynamics simulation, closed-loop evaluation, and safety-critical scenario testing.
• Safety assurance, verification and validation, formal methods, scenario generation, digital twins, simulation-to-reality transfer, and field testing;
• Real-time implementation, embedded and edge computing, resource-efficient algorithms, energy-aware computation, and deployment on production-relevant platforms;
• Human factors, social compliance, comfort, acceptance, datasets, benchmarks, evaluation protocols, open-source platforms, and reproducibility.

We particularly encourage contributions that provide transparent comparisons between methodological families and demonstrate validated improvements in safety, robustness, efficiency, comfort, generalization, or real-world deployability.

Dr. Yongqi Dong
Dr. Xiaoxi Hu
Dr. Yiyun Wang
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-anonymized peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Vehicles is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.