State classification for autonomous gas sample taking using deep convolutional neural networks

Despite recent rapid advances and successful large-scale application of deep Convolutional Neural Networks (CNNs) using image, video, sound, text and time-series data, its adoption within the oil and gas industry in particular have been sparse. In this paper, we initially present an overview of opportunities for deep CNN methods within oil and gas industry, followed by details on a novel development where deep CNN have been used for state classification of autonomous gas sample taking procedure utilizing an industrial robot. The experimental results - using a deep CNN containing six layers - show accuracy levels exceeding 99 %. In addition, the advantages of using parallel computing with GPU is re-confirmed by showing a reduction factor of 43,8 for the training time required as compared with a CPU implementation. Finally, by analyzing the variations in the output probability distribution, it is shown that the deep CNN can also detect a number of undefined and therefore untrained anomalies. This is an extremely appealing property and serves as an illustrative example of how deep CNN algorithms can contribute towards safer and more robust operation in the industry.