Robust Automated Driving in Extreme Weather

Informazioni relative al progetto

ROADVIEW

ID dell’accordo di sovvenzione: 101069576

DOI

10.3030/101069576

Data della firma CE 9 Giugno 2022

Data di avvio 1 Settembre 2022

Data di completamento 31 Agosto 2026

Finanziato da

• Climate, Energy and Mobility

Costo totale

€ 8 014 105,55

Contributo UE

€ 6 652 915,80

Coordinato da

HOGSKOLAN I HALMSTAD
Sweden

CORDIS fornisce collegamenti ai risultati finali pubblici e alle pubblicazioni dei progetti ORIZZONTE.

I link ai risultati e alle pubblicazioni dei progetti del 7° PQ, così come i link ad alcuni tipi di risultati specifici come dataset e software, sono recuperati dinamicamente da .OpenAIRE .

Risultati finali

Documents, reports (23)

SW on Improved Localization Using High-density Map Updating - Final report

A final report related SW of a compressed HD map representation that can cope with the seasonal changes and can be kept updated while excluding the dynamic objects.

Final KPI specifications

Update on the KPI specifications.

Evaluation of sensors in-situ

Report on robustness of shielded sensors based on data while driving in real-world harsh weather conditions, including possible mitigations such as design choices, filtering, and calibration.

Test plan regarding the most appropriate test method

Report defining how to test single components and integrated automated vehicle systems in virtual and real domains. It will consider specific uses-cases under extreme weather conditions.

Report on data privacy and storage

Data privacy and storage of the logged data.

Final readiness assessment of specific datasets

Final report on the quantitative evaluation of the ROADVIEW DRL method on public datasets and on the ROADVIEW dataset using enlarged feature sets.

Initial readiness assessment of specific datasets

Report on the quantitative evaluation of the ROADVIEW DRL method on public datasets using a limited feature size.

Interim impact report

Interim impact report on communication and dissemination activities, including those at the international level. An update of the DCP plan will be considered.

SW on Improved Localization Using High-density Map Updating - First report

A first report related SW of a compressed HD map representation that can cope with the seasonal changes and can be kept updated while excluding the dynamic objects.

SW on Collaborative Perception Solutions - Final report

A final report on improving vehicle perception with sharing information provided by sensors from other connected vehicles or VRUs and roadside infrastructure.

SW on Adaptive Sensor Fusion and Perception Solutions - Final report

A final report on algorithms and SW developed to detect weather-related physical conditions (e.g., heavy rain, fog, slush and snow), free space detection, object detection and tracking, and adaptive sensor fusion.

SW on Adaptive Sensor Fusion and Perception Solutions - First report

A first report on algorithms and SW developed to detect weather-related physical conditions (e.g., heavy rain, fog, slush and snow), free space detection, object detection and tracking, and adaptive sensor fusion.

Stakeholder report and strategy

Stakeholder report and strategy including detailed stakeholder mapping and strategies to reach key target audiences.

Environmental and weather condition state estimates

Report on the online state estimation method providing the reliable visibility range and weather conditions using camera, LiDAR, and RADAR sensor data.

Vehicle dynamics modelling methodology

A public report will be published along with any relevant publications on vehicle dynamics modelling methodologies. In addition, a representative library of vehicle dynamics models for a selection of vehicles available within the consortium will be delivered.

Definition of the complex environment conditions

An extended ODD taxonomy will be reported considering harsh weather conditions and complex urban/rural environments.

XAI design of perception solutions

Explainability requirement specifications, report and implementation of XAI design for AI models developed in Tasks 5.2 and 5.3.

SW on Collaborative Perception Solutions - First report

A first report on improving vehicle perception with sharing information provided by sensors from other connected vehicles or VRUs and roadside infrastructure.

Report on data annotation

Report on the categorisation and labelling of data.

Use cases and scenarios

A set of use cases with multiple scenarios will be reported based on the ODD taxonomy defined in Task 2.1.

Plan for the dissemination and communication activities

Plan for the dissemination and communication activities including defined strategy, tools, channels and (international) activities.

Initial KPI specifications

Report on the KPI specifications.

Relevant metrics to assert the validity of the simulation-assisted test methods

This report describes the relevant metrics for the validation of simulation-assisted methods in comparison to controlled environment/proving ground reference values.

Other (3)

Library of internal sensor noise models and compound noise modelling methodology

A library of internal sensor noise models for the perception sensors selected for the project will be delivered. This will be accompanied by a report on noise modelling, the methodologies to combine multiple noise factor effects on the sensor outputs for different sensor technologies, and any significant corner cases or scenarios to test where compound noise could create significant degradation of sensor performance.

Library of validated physics-based parameterised noise models

A library of validated physics-based parameterised noise models for the selected perception sensor technologies with detailed report on implementation, validation methods and comparison of the generated models based on the agreed evaluation metrics.

Library of validated statistical noise models

A library of validated statistical noise models for the selected perception sensor technologies will be delivered with a detailed report on implementation and validation methods used.

Data sets, microdata, etc (2)

GitHub release of ROADVIEW Readiness levels

Release of the source code for DRL.

Reference dataset of measured weather characteristics

Reference datasets for the controlled environment and real-world site weather characteristics will be delivered, with a detailed report on data collection methodology.

Websites, patent fillings, videos etc. (1)

ROADVIEW website

ROADVIEW website with dedicated areas for different stakeholder groups.

Demonstrators, pilots, prototypes (1)

ROADVIEW Demonstration 1

Report on the demonstration of early ROADVIEW system components implementations to be used in the subsequent demonstrations.

Pubblicazioni

Conference proceedings (9)

Synthetic Extreme Weather for AI training: Concept and Validation

Autori: Letícia Cristófoli Duarte Silva, Maikol Funk Drechsler, Yuri Poledna, Werner Huber, Thiago Antonio Fiorentin
Pubblicato in: 2023 Third International Conference on Digital Data Processing (DDP), 2024, ISSN 2473-2001
Editore: IEEExplore
DOI: 10.1109/DDP60485.2023.00044

Creation of digital models for accelerated and reliable testing of automated systems in adverse weather

Autori: Tuomas Herranen, Erik Henriksson, Pak Hung Chan, Yuri Poledna, Pierre Duthon, Amine Ben-Daoued, Maikol Drechsler, Valentina Donzella
Pubblicato in: Autonomous Systems for Security and Defence, 2024
Editore: SPIE
DOI: 10.1117/12.3031473

Semantics-aware LiDAR-Only Pseudo Point Cloud Generation for 3D Object Detection

Autori: Tiago Cortinhal, Idriss Gouigah, Eren Erdal Aksoy
Pubblicato in: 2024 IEEE Intelligent Vehicles Symposium (IV), 2024
Editore: IEEE
DOI: 10.1109/IV55156.2024.10588782

Vehicle Dynamics Parameter Estimation Methodology for Virtual Automated Driving Testing

Autori: Maikol Funk Drechsler, Yuri Poledna, Mattias Hjort, Sogol Kharrazi, Werner Huber
Pubblicato in: 2024 IEEE International Automated Vehicle Validation Conference (IAVVC), 2024
Editore: IEEE
DOI: 10.1109/IAVVC63304.2024.10786416

REHEARSE: adveRse wEatHEr datAset for sensoRy noiSe modEls

Autori: Yuri Poledna, Maikol Funk Drechsler, Valentina Donzella, Pak Hung Chan, Pierre Duthon, Werner Huber
Pubblicato in: 2024 IEEE Intelligent Vehicles Symposium (IV), 2024
Editore: IEEE
DOI: 10.1109/IV55156.2024.10588491

3D-OutDet: A Fast and Memory Efficient Outlier Detector for 3D LiDAR Point Clouds in Adverse Weather

Autori: Raisuddin, Abu Mohammed; Cortinhal, Tiago; Holmblad, Jesper; Aksoy, Eren Erdal
Pubblicato in: 2024 IEEE Intelligent Vehicles Symposium (IV), 2024
Editore: IEEE
DOI: 10.36227/TECHRXIV.24297166.V1

Simulation numérique de capteurs perceptifs du véhicule autonome sous conditions météorologiques dégradées

Autori: Amine Ben-Daoued; Frédéric Bernardin; Pierre Duthon
Pubblicato in: ATEC ITS Congress, 2023
Editore: HAL

The effect of camera data degradation factors on panoptic segmentation for automated driving

Autori: Wang, Yiting, Zhao, Haonan, Debattista, Kurt and Donzella, Valentina
Pubblicato in: 26th IEEE International Conference on Intelligent Transportation Systems (ITSC 2023), 2023
Editore: WRAP Warwick

Parametric Physics-Based Snow Model for Automotive Cameras

Autori: Pak Hung Chan, Kurt Debattista, Valentina Donzella
Pubblicato in: 2024 IEEE SENSORS, 2024
Editore: IEEE
DOI: 10.1109/SENSORS60989.2024.10784836

Peer reviewed articles (11)

A Noise Analysis of 4D RADAR: Robust Sensing for Automotive?

Autori: Pak Hung Chan; Sepeedeh Shahbeigi Roudposhti; Xinyi Ye; Valentina Donzella
Pubblicato in: IEEE Sensors Journal, 2025, ISSN 1558-1748
Editore: IEEE
DOI: 10.36227/TECHRXIV.24517249

Raw Camera Data Object Detectors: An Optimisation for Automotive Video Processing and Transmission

Autori: Pak Hung Chan, Chuheng Wei, Anthony Huggett, Valentina Donzella
Pubblicato in: IEEE Access, Numero 13, 2025, ISSN 2169-3536
Editore: IEEE
DOI: 10.36227/TECHRXIV.23807499.V1

The inconvenient truth of ground truth errors in automotive datasets and DNN-based detection

Autori: Chan, Pak Hung; Li, Boda; Baris, Gabriele; Sadiq, Qasim; Donzella, Valentina
Pubblicato in: Data-Centric Engineering, 2024, ISSN 2632-6736
Editore: Cambridge University Press
DOI: 10.36227/TECHRXIV.170630689.97107789/V1

A Novel Score-based LiDAR Point Cloud degradation Analysis Method

Autori: Sepeedeh Shahbeigi, Honahan Robinson, Valentina Donzella
Pubblicato in: IEEE Transactions and Journals, 2024, ISSN 1803-7232
Editore: IEEE Xplore
DOI: 10.1109/ACCESS.2024.3359300

Automotive DNN-Based Object Detection in the Presence of Lens Obstruction and Video Compression

Autori: Gabriele Baris, Boda Li, Pak Hung Chan, Carlo Alberto Avizzano, Valentina Donzella
Pubblicato in: IEEE Access, Numero 13, 2025, ISSN 2169-3536
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.1109/ACCESS.2025.3544773

A Comparative Review of the SWEET Simulator: Theoretical Verification Against Other Simulators

Autori: Amine Ben-Daoued; Frédéric Bernardin; Pierre Duthon
Pubblicato in: Journal of Imaging, 2024, ISSN 3064-6936
Editore: MDPI
DOI: 10.3390/JIMAGING10120306

LiDAR De-Snow Score (DSS): Combining Quality and Perception Metrics for Optimized De-Noising

Autori: Valentina Donzella, Pak Hung Chan, Daniel Gummadi, Abu Mohammed Raisuddin, Eren Erdal Aksoy
Pubblicato in: IEEE Sensors Journal, Numero 25, 2025, ISSN 1530-437X
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.1109/JSEN.2025.3570478

Depth- and semantics-aware multi-modal domain translation: Generating 3D panoramic color images from LiDAR point clouds

Autori: Tiago Cortinhal, Eren Erdal Aksoy
Pubblicato in: Robotics and Autonomous Systems, 2023, ISSN 0921-8890
Editore: Science Direct | Elsevier
DOI: 10.1016/j.robot.2023.104583

SWEET: A Realistic Multiwavelength 3D Simulator for Automotive Perceptive Sensors in Foggy Conditions

Autori: Amine Ben-Daoued; Pierre Duthon; Frédéric Bernardin
Pubblicato in: Journal of Imaging, 2023, ISSN 2313-433X
Editore: MDPI
DOI: 10.3390/jimaging9020054

Influence of AVC and HEVC Compression on Detection of Vehicles Through Faster R-CNN

Autori: Pak Hung Chan, Anthony Huggett, Georgina Souvalioti, Paul Jennings, Valentina Donzella
Pubblicato in: IEEE Transactions on Intelligent Transportation Systems, Numero 25, 2025, ISSN 1524-9050
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.36227/TECHRXIV.19808566

From operational design domain to test cases: A methodology to include harsh weather

Autori: Fredrik Warg; Valentina Donzella; Pak Hung Chan; Jonathan Robinson; Yuri Poledna; Sebastien Liandrat; Umut Cihan; Maytheewat Aramrattana; Graham Lee; Eren Erdal Aksoy
Pubblicato in: Open Research Europe, 2024, ISSN 2732-5121
Editore: European Commission
DOI: 10.12688/OPENRESEUROPE.18592.1

Other (7)

3D-OutDet: A Fast and Memory Efficient Outlier Detector for 3D LiDAR Point Clouds in Adverse Weather

Autori: Abu Mohammed Raisuddin, Tiago Cortinhal, Jesper Holmblad, Eren Erdal Aksoy
Pubblicato in: TechRxiv, 2023
Editore: TechRxiv
DOI: 10.36227/techrxiv.24297166.v1

Semantics-aware LiDAR-Only Pseudo Point Cloud Generation for 3D Object Detection

Autori: Tiago Cortinhal, Idriss Gouigah, Eren Erdal Aksoy
Pubblicato in: arXiv, 2023
Editore: arXiv
DOI: 10.48550/arXiv.2309.08932

Raw camera data object detectors: an optimisation for automotive processing and transmission

Autori: Pak Hung Chan, Chuheng Wei, Anthony Huggett, Valentina Donzella
Pubblicato in: TechRxiv, 2023
Editore: TechRxiv
DOI: 10.36227/techrxiv.23807499.v1

Correlating traditional image quality metrics and DNN-based object detection: a case study with compressed camera data

Autori: Daniel Gummadi, Pak Hung Chan, Hetian Wang, Valentina Donzella
Pubblicato in: TechRxiv, 2023
Editore: TechRxiv
DOI: 10.36227/techrxiv.24566371.v1

Analysis of Faster R-CNN network prediction in the presence of lens occlusion and video compression

Autori: Gabriele Baris, Boda Li, Pak Hung Chan, Carlo Alberto Avizzano, Valentina Donzella
Pubblicato in: TechRxiv, 2023
Editore: TechRxiv
DOI: 10.36227/techrxiv.23047412.v1

An Experimental Study on ObjectTracking

Autori: Mahmoud Alshaikh
Pubblicato in: 2025
Editore: n/a

A noise analysis of 4D RADAR: robust sensing for automotive?

Autori: Pak Hung Chan, Sepeedeh Shahbeigi Roudposhti, Xinyi Ye, Valentina Donzella
Pubblicato in: TechRxiv, 2023
Editore: TechRxiv
DOI: 10.36227/techrxiv.24517249.v1

Thesis and dissertations (1)

Pixelwise Road Surface Slipperiness Estimation for Autonomous Driving with Weakly Supervised Learning

Autori: Julius Pesonen
Pubblicato in: Machine Learning, Data Science and Artificial Intelligence, 2023
Editore: Aaltodoc publication archive

Book chapters (2)

Road Grip Uncertainty Estimation Through Surface State Segmentation

Autori: Jyri Maanpää, Julius Pesonen, Iaroslav Melekhov, Heikki Hyyti, Juha Hyyppä
Pubblicato in: Lecture Notes in Computer Science, Image Analysis, 2025
Editore: Springer Nature Switzerland
DOI: 10.1007/978-3-031-95911-0_17

Dense Road Surface Grip Map Prediction from Multimodal Image Data

Autori: Jyri Maanpää, Julius Pesonen, Heikki Hyyti, Iaroslav Melekhov, Juho Kannala, Petri Manninen, Antero Kukko, Juha Hyyppä
Pubblicato in: Lecture Notes in Computer Science, Pattern Recognition, 2024
Editore: Springer Nature Switzerland
DOI: 10.1007/978-3-031-78447-7_26