Impact
Compartir
Save the reference
Title
A Probabilistic Physics-Aware Battery Health Management Approach for Inspection Drone Operations
Other institutions
https://ror.org/000xsnr85Version
PostprintDocument type
Book PartLanguage
EnglishRights
© 2026, The Author(s), under exclusive license to Springer Nature Switzerland AGAccess
Open accessIdentificador
https://doi.org/10.1007/978-3-032-21502-4_11Published at
Digital Maintenance and Asset Digitalization Engineering Asset Management Review. Vol 5.Publisher
Springer NatureKeywords
Probabilistic battery health management
Physics-aware machine learning
Uncertainty quantification
End-of-discharge voltage prediction ... [+]
Physics-aware machine learning
Uncertainty quantification
End-of-discharge voltage prediction ... [+]
Probabilistic battery health management
Physics-aware machine learning
Uncertainty quantification
End-of-discharge voltage prediction
Drone-based infrastructure inspection
ODS 7 Energía asequible y no contaminante [-]
Physics-aware machine learning
Uncertainty quantification
End-of-discharge voltage prediction
Drone-based infrastructure inspection
ODS 7 Energía asequible y no contaminante [-]
Subject (UNESCO Thesaurus)
Data analysisData protection
Abstract
The increasing deployment of inspection drones for monitoring remote and critical infrastructure presents new opportunities and challenges in asset management. These drones operate in demanding enviro ... [+]
The increasing deployment of inspection drones for monitoring remote and critical infrastructure presents new opportunities and challenges in asset management. These drones operate in demanding environments, where ensuring operational reliability is essential. Among the various subsystems, battery health plays a central role in determining mission success and safety. This chapter presents a physics-aware probabilistic approach for battery health management, integrating data-driven techniques with physics-based models to improve the predictability of battery performance. At the core of this methodology lies a probabilistic machine learning model that provides uncertainty quantification, enabling more informed and robust decision-making. By incorporating this uncertainty-aware perspective into battery discharge forecasting, the approach supports advanced digital maintenance strategies. The methodology is demonstrated to drone-based inspections of offshore wind energy infrastructure, highlighting its contribution to enhancing asset reliability and enabling condition-aware maintenance strategies. [-]
