Mycelium-based biocomposites (MBBs) represent a sustainable alternative to synthetic composites, as they are produced from lignocellulosic substrates bonded by fungal mycelium. Their mechanical performance depends on multiple interacting factors, including the substrate composition, fungal species, and processing conditions, which makes property optimisation challenging. In this study, an artificial neural network (ANN) model was developed to predict two mechanical properties of MBBs, namely internal bonding (IB) and compressive strength (CS). An ANN model was trained on experimental data, using the substrate composition, fungal species, and physical properties of MBBs. The ANN predictions were compared with measured values, and the model accuracy was evaluated. The results showed that the ANN achieved a high predictive accuracy, with coefficients of determination of 0.992 for IB and 0.979 for CS. IB values were predicted more precisely than CS, likely due to microstructural heterogeneities. The heterogeneities were visualised using scanning electron microscopy. Composites produced with Ganoderma sessile and Trametes versicolor exhibited the highest IB. Interestingly, Trametes versicolor achieved the highest CS on virgin wood particles but the lowest values on recycled wood, underlining the strong influence of the substrate quality. The study demonstrates that ANNs can effectively predict the mechanical properties, reducing the number of experimental tests needed for material characterisation.
Hýsek et al. (2025) studied this question.