DIGITAL TWIN PLATFORM FOR AGRO-WASTE BIODEGRADABLE PACKAGING IN CENTRAL ASIA: IOT QUALITY AND ML RECIPE OPTIMISATION

Abstract

The transition from petroleum-based packaging to biodegradable alternatives is constrained by feedstock variability, slow recipe development, and limited process controllability at pilot scale. This paper proposes an integrated Industry 4.0 framework for producing biodegradable packaging biocomposites from locally available agro-waste (e.g., cereal straw and rice husk) through the BioPack Core module embedded in the R&D Central Asia platform. The approach combines (i) an IoT-enabled data layer for raw-material traceability and in-line monitoring, (ii) a production-line digital twin for predictive process optimisation and virtual commissioning, and (iii) machine-learning models for formulation screening and multi-objective recipe optimisation. We present a methodological pipeline that links agro-waste physicochemical descriptors with mechanical, barrier, and biodegradation performance targets, enabling accelerated down-selection of candidate biocomposites and risk-reduced scale-up. The proposed architecture contributes a reproducible digital thread from «waste-to-recipe-to-product» supporting quality management, faster R&D cycles, and scalable circular bioeconomy deployment in Central Asia