In this case study, we present a prototypical application of anomaly detection within an industrial setting, utilizing a bespoke toy dataset. This section features a brief demonstration of a CNN composed of three layers, followed by a fully connected (FC) layer, tasked with performing binary classification for anomaly detection. This implementation is carried out using the PyTorch deep learning framework (//pytorch.org/). Our dataset is composed of top-view photographs of a singular, yellow toy car. For the purposes of this study, an anomaly is defined as the car having one or both of its doors open.