Perception Enhancement and Improving Driving Context Recognition of an Autonomous Vehicle Using UAVs

Khezaz, Abderraouf; Hina, Manolo Dulva; Ramdane-Cherif, Amar · 2022 · Crossref

DOI: 10.3390/jsan11040056

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the critical challenge of enhancing environmental perception for autonomous vehicles, a prerequisite for safe navigation and decision-making. The authors argue that relying solely on onboard sensors is insufficient due to limitations in field of view, weather dependency, and potential sensor faults. To mitigate these risks, the study proposes integrating Unmanned Aerial Vehicles (UAVs) as external sensory tools. UAVs, equipped with radar, lidar, and cameras, can gather data from elevated positions and transmit it to the vehicle, thereby extending the perception range and providing redundancy. The motivation stems from the need for a robust, dependable perception system that functions reliably even when communication grids are compromised or visibility is poor. The methodology employs a knowledge-based system centered on an ontology to represent the driving context. Data from both local vehicle sensors and external UAVs are fused within this ontology, which acts as a knowledge base containing entities, properties, and logical rules. The system utilizes Semantic Web Rule Language (SWRL) for reasoning and fuzzy logic to handle uncertainty in sensor reliability and environmental conditions. Specifically, fuzzy membership functions quantify variables such as weather, brightness, and sensor quality into crisp values (e.g., "Good," "Average," "Poor") to determine an overall perception score. The reasoner uses these scores and predefined logical rules to decide when to activate UAVs and which communication protocols—Radio Frequency (RF), Visible Light Communication (VLC), or a hybrid approach—to use for secure data transmission. As a proof of concept, the proposed system was tested using various use cases in a laboratory driving simulator. The experimental design validated the system's ability to process multimodal data, manage sensor uncertainties via fuzzy logic, and execute logical rules for UAV deployment and communication selection. The results demonstrated that the integration of UAVs significantly improves driving context recognition. The system effectively combined local and external perception data to provide a more accurate and reliable understanding of the driving environment. The significance of this work lies in its demonstration that UAVs can serve as effective dynamic data-gathering tools for intelligent transportation systems. By enhancing perception accuracy, the system contributes to safer autonomous driving and accident prevention. The use of ontology and fuzzy logic provides a scalable and rigorous framework for managing complex, uncertain data from multiple sources. This approach offers a viable solution for overcoming the limitations of traditional vehicular sensors, particularly in adverse weather or complex urban environments, thereby advancing the reliability of autonomous navigation processes.

Key finding

The integration of UAV-collected data into an ontology-based knowledge base significantly enhances autonomous vehicle perception and driving context recognition, as validated through simulator experiments.

Methodology

simulator

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• situational awareness