🧪 🔬Presented work

Paper 1. “Adaptive Parameter Aggregation in Federated Learning for Synchronizing Digital Twins of Manufacturing Processes Under Seasonality and Data Drift”
Authors: Gulshat Amirkhanova, Liya Bektemir, Tomiris Nurgazy, Alikhan Amirkhanov
Summary: a method of adaptive model parameter aggregation in federated learning for synchronizing the digital twins of distributed manufacturing systems. The algorithm adjusts the weights of local models using three coefficients — seasonal correlation, statistical distribution drift and edge node reliability. On open IIoT datasets, the method improved the anomaly detection F1-score by 22% over the FedAvg baseline under conditions of seasonality and drift.

Paper 2. “Cognitive Armor: AI-Powered Communication Security for the Digital Twin Era”
Authors: Saltanat Adilzhanova, Gulshat Amirkhanova, Bauyrzhan Amirkhanov, Dana Sybanova, Murat Kunelbayev
Summary: a unified AI security framework that ensures data integrity, authenticity and resilience across heterogeneous environments — from industrial digital twin ecosystems to corporate email. Within the industrial perimeter, it combines machine learning, natural language processing (NLP) and behavioural analytics to protect data exchange between suppliers, manufacturers and quality control systems, including supply chain verification and HACCP compliance monitoring.