The use of artificial intelligence (AI) in waste management has enormous benefits for the entire ecosystem and additionally diminishes the stress on the public health system given the substantial amount of waste being produced today. This project addresses this important social problem of solid waste segregation where an attempt has been made to classify plastics and non-plastics using SSD-MobilenetV2. SSD-MobileNetV2 uses depthwise convolution and pointwise convolution that make the various channels smaller and the residual block, decreases the amount of data on the network, making detection times faster than other models. The SSD model is trained on a custom dataset to obtain an efficient a deep learning model that can retain the spatial organisation of the input image at a lesser resolution while extracting semantic meaning from it. The research focuses on boosting the performance of pretrained models on a custom dataset and in this work, the deep learning MobileNetV2 is incorporated into SSD to achieve efficient and fast detection.