Predictive Modelling of Silver and Thickness Measurements in Medical Foam Production - Interpretable Machine Learning for Process Understanding

Sammanfattning

Medical foam used in wound-care products must meet quality requirements for propertiessuch as thickness and silver content. This thesis investigates whether routinelycollected production data can be used to predict these properties and to supportprocess understanding. The study uses six months of data from medical foam production,including quality-control and laboratory measurements, roll information,and process sensor data from the production equipment.The raw data sources were cleaned, linked to finished production rolls or roll positions,and converted into modelling tables for four responses: silver concentration,Ag [mg/g]; silver area-density, Ag [mg/cm2]; average roll thickness; and thicknesslog variance, which describes variation in thickness within a roll. Several regressionmodels were compared using time-based cross-validation and a held-out laterproduction period, so that model performance was evaluated in a setting similar topredicting future production.The results depended strongly on the response. Silver concentration was not usefullypredictable from the available data, with test R2 = 0.005. Silver area-densityshowed some predictive signal during the earlier development period, but the modelsdid not generalise to the later test period because silver levels shifted upward. Incontrast, average thickness was predicted very accurately from process sensor summariesalone, with the best model reaching test R2 = 0.989. Thickness log variancewas harder to predict but still showed useful signal, with the best model reachingtest R2 = 0.652. The most stable interpretation results therefore concern thickness.The fitted models identified production contexts and sensor summaries thatcan guide expert process review, especially for thickness level and foam uniformity.These results should be read as predictive associations, not as causal evidence ordirect control rules.