RESUMO
Technological innovations are not enough by themselves to achieve social and environmental sustainability in companies. Sustainable development aims to determine the environmental impact of a product and the hidden price of products and services through the concept of radical transparency. This means that companies should show and disclose the impact on the environment of any good or service. This way, the consumer can choose in a transparent manner, not only for the price. The use of the eco-label as a European eco-label, which bases its criteria on life cycle assessment, could provide an indicator of corporate social responsibility for a given product. However, it does not give a full guarantee that the product was obtained in a sustainable manner. The aim of this work is to provide a way of calculating the value of the environmental impacts of an industrial product, under different operating conditions, so that each company can provide detailed information on the impacts of its products, information that can form part of its "green product sheet". As a case study, the daily production of a newspaper, printed by coldset, has been chosen. Each process involved in production was configured with raw material and energy consumption information from production plants, manufacturer data and existing databases. Four non-linear regression models have been trained to estimate the impact of a newspaper's circulation from five input variables (pages, grammage, height, paper type, and print run) with 5508 data samples each. These non-linear regression models were trained using the Levenberg-Marquardt nonlinear least squares algorithm. The mean absolute percentage errors (MAPE) obtained by all the non-linear regression models tested were less than 5%. Through the proposed correlations, it is possible to obtain a score that reports on the impact of the product for different operating conditions and several types of raw materials. Ecolabelling can be further developed by incorporating a scoring system for the impact caused by the product or process, using a standardised impact methodology.
RESUMO
This research study focuses on the evaluation of the total phenolic compounds (TPC) and antioxidant activity (AOA) of strawberries according to different experimental extraction conditions by applying the Artificial Neural Networks (ANNs) technique. The experimental data were applied to train ANNs using feed- and cascade-forward backpropagation models with Levenberg-Marquardt (LM) and Bayesian Regulation (BR) algorithms. Three independent variables (solvent concentration, volume/mass ratio and extraction time) were used as ANN inputs, whereas the three variables of total phenolic compounds, DPPH and ABTS antioxidant activities were considered as ANN outputs. The results demonstrate that the best cascade- and feed-forward backpropagation topologies of ANNs for the prediction of total phenolic compounds and DPPH and ABTS antioxidant activity factors were the 3-9-1, 3-4-4-1 and 3-13-10-1 structures, with the training algorithms of trainlm, trainbr, trainlm and threshold functions of tansig-purelin, tansig-tansig-tansig and purelin-tansig-tansig, respectively. The best R2 values for the predication of total phenolic compounds and DPPH and ABTS antioxidant activity factors were 0.9806 (MSE = 0.0047), 0.9651 (MSE = 0.0035) and 0.9756 (MSE = 0.00286), respectively. According to the comparison of ANNs, the results showed that the cascade-forward backpropagation network showed better performance than the feed-forward backpropagation network for predicting the TPC, and the FFBP network, in predicting the DPPH and ABTS antioxidant activity factors, had more precision than the cascade-forward backpropagation network. The ANN technique is a potential method for estimating targeted total phenolic compounds and the antioxidant activity of strawberries.