ANSR-DT: A Neuro-Symbolic Framework for Adaptive and Explainable Digital Twins

Researchers have proposed two new frameworks, ANSR-DT and HAPI, aimed at enhancing the capabilities of digital twins in industrial and medical applications.

The ANSR-DT framework, detailed in a paper submitted to arXiv on 15 Jan 2025^[1], combines a CNN-LSTM model with Prolog-based reasoning to enable adaptive and explainable digital twins. It incorporates a PPO-based adaptation layer for refining operational responses. ANSR-DT was compared against 8 baselines and demonstrated competitive predictive performance and stable rule extraction. The framework's development involved multiple submissions, with the first version being 2,469 KB in size in 2025^[1] and the fourth version being 845 KB in size in 2026^[1]. Meanwhile, a separate research effort introduced HAPI, a framework designed for patient-specific digital twins, such as a heart model, which was submitted on 14 Jun 2026^[2]. HAPI constructs a physics-integrated gray-box model with an interpretable mechanistic backbone and a neural component, enabling rapid adaptation to live data. The HAPI framework's submission was 20,999 KB in size^[2]. Both ANSR-DT and HAPI aim to address limitations in existing digital twin frameworks, such as interpretability and adaptability.