A Continual Validation, Updating, and Decision-Making Framework for Self-Adaptive Digital Twins via Robust Model Predictive Control

Adaptive Digital Twin framework uses LoRA and drift detection for continual surrogate model validation

A new arXiv preprint presents a framework for self-adaptive Digital Twins that combines Fisher score-based drift detection, Low-Rank Adaptation (LoRA) for parameter-efficient continual learning, and Mann-Whitney U tests for online statistical validation of surrogate model updates. The system triggers fine-tuning of fewer than 1% of model parameters upon drift detection and certifies predictive improvement before deploying updated surrogates. Case studies in stochastic linear systems and directed energy deposition additive manufacturing demonstrate successful detection of distributional shifts with short delays. The work addresses the open challenge of maintaining surrogate model fidelity under concept drift in real-time physical system mirroring.