SmartNanoSense Innovation use specialist software for model-based design and simulation, such as MathWorks Matlab and Simulink, dSPACE for model-based design, and COMSOL Multiphysics for multiphase flow simulation. SmartNanoSense Innovation also employs a dedicated compute cluster for machine learning, neural network training and continuous verification. This includes classification models, following Machine Learning (ML) and Deep Learning (DL) approaches for the various sensors and classification problems, such as DL Convolutional Neural Networks (CNNs) with a final block of Fully Connected (FC) layers can be used to directly classify signals into categories. When dealing with longitudinal signal samples (time-series of signals from longitudinal tests), more complex DL schemes are used to better handle temporal relationships and recurrence. These include CNNs followed by a Recurrent Neural Network (RNN) block, Quasi-Recurrent Neural Networks (QNNs), and the recent Transformers Networks.