Tools for monitoring toxicological and genotoxicological changes in a drinking water treatment plant in Northeast Italy

Drinking water quality can be compromised at different stages, from raw water to treated one. This research aimed to evaluate the toxicity and genotoxicity of groundwater contaminated by fluorinated compounds treated in a drinking water treatment plant, through several bioassays. Water samples underwent chemical analyses and were assayed on Daphnia magna, Pseudokirchneriella subcapitata, Allium cepa, human leukocytes and Salmonella typhimurium. Physical–chemical parameters were always within the Italian legislation limits. Water after filtration and disinfection caused slight toxicity in D. magna; the sample after filtration inhibited the proliferation of P. subcapitata. None of the water samples exerted toxicity in A. cepa. All the analysed samples had genotoxic effects on A. cepa and human leucocytes, while only disinfected water caused mutations in S. typhimurium. A battery composed of tests on D. magna, P. subcapitata, S. typhimurium and A. cepa could represent a useful tool to verify the toxicity/genotoxicity through the water treatment stages and to improve drinking water quality management.     

Autoencoders for unsupervised real-time bridge health assessment

Over the last decades, the rising number of aging infrastructures has progressively fueled much interest toward the field of structural health monitoring. Following the increasing popularity of artificial intelligence algorithms, an autoencoder-based damage detection technique within the context of unsupervised learning is proposed in this paper to provide support for practical engineering applications. The developed methodology uses the autoencoder to reconstruct raw acceleration sequences of user-defined length collected from a healthy structure. To quantify the errors between the original input and the reconstructed output, which may be representative of damage occurrence, two indexes of reconstruction loss are selected as damage-sensitive features. To support damage detection, a selected number of short-time sequences are finally grouped into a unique macrosequence. The novel procedure can effectively both work at the single sensor level, as well as combine the predictive models using an ensemble learning strategy. Avoiding system identification, results obtained in the Z24 bridge demonstrate that the proposed method is quite effective for local damage detection with limited computational effort and using a limited number of sensors, thereby suitable to be easily applicable in the context of real-time bridge assessment.