EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

TrustLife: A Trust-Aware Artificial Intelligence Lifecycle Management Framework for Secure Embodied Robots with Multi-Cloud Deployment Optimization

Linting Wu, Jinxiang Yu

Fundamental Scientific Reports in Multidisciplinary Areas · 27.10.2026 · 2 (3) · 271-286

DOI: 10.66238/fsrma140 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Fundamental Scientific Reports in Multidisciplinary Areas, online erschienen am 27.10.2026. Verfasst von Linting Wu und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Embodied Artificial Intelligence (Embodied AI) is becoming a key enabling technology for intelligent robots in manufacturing, healthcare, logistics, and service environments, where continuous model evolution and secure cloud-based deployment are essential for reliable human–robot collaboration. However, existing Machine Learning Operations (MLOps) frameworks primarily emphasize model training and deployment efficiency while overlooking trustworthy lifecycle management, safety validation, and adaptive deployment across heterogeneous multi-cloud infrastructures. To address these challenges, this paper proposes TrustLife, a Trust-Aware Artificial Intelligence (AI) Lifecycle Management Framework for secure embodied robots with multi-cloud deployment optimization. TrustLife integrates Dynamic Quality Assessment (DQA), Adaptive Automated Machine Learning (Adaptive AutoML), a Trust Verification Engine (TVE), a Multi-Cloud Deployment Optimizer (MCDO), runtime monitoring, and adaptive retraining into a unified closed-loop architecture that continuously manages the entire AI lifecycle from data governance to online model evolution. By jointly optimizing model trustworthiness, deployment latency, computational resource allocation, operational cost, and runtime reliability, the proposed framework enables secure and adaptive deployment across heterogeneous cloud platforms. Experimental evaluations on embodied robotics benchmarks and heterogeneous multi-cloud environments demonstrate that TrustLife achieves a deployment success rate of 97.2%, a runtime reliability of 96.7%, and an average inference latency of 129 ms, consistently outperforming state-of-the-art AI lifecycle management frameworks. These results verify that TrustLife effectively enhances trustworthy AI lifecycle management while improving deployment efficiency, operational stability, and cloud resource utilization for secure embodied robotic systems.

Zitieren

APA

Wu, L., & Yu, J. (2026). TrustLife: A Trust-Aware Artificial Intelligence Lifecycle Management Framework for Secure Embodied Robots with Multi-Cloud Deployment Optimization. Fundamental Scientific Reports in Multidisciplinary Areas, 2(3), 271-286. https://doi.org/10.66238/fsrma140

Vancouver

Wu L, Yu J. TrustLife: A Trust-Aware Artificial Intelligence Lifecycle Management Framework for Secure Embodied Robots with Multi-Cloud Deployment Optimization. Fundamental Scientific Reports in Multidisciplinary Areas. 2026;2(3):271-286. doi:10.66238/fsrma140

BibTeX
@article{Wu2026,
  author = {Wu, Linting and Yu, Jinxiang},
  title = {TrustLife: A Trust-Aware Artificial Intelligence Lifecycle Management Framework for Secure Embodied Robots with Multi-Cloud Deployment Optimization},
  journal = {Fundamental Scientific Reports in Multidisciplinary Areas},
  year = {2026},
  volume = {2},
  number = {3},
  pages = {271--286},
  doi = {10.66238/fsrma140}
}
Kostenlos starten →