This research presents a model for detecting pipeline vandalism in oil and gas sector. Feed-forward deep learning technique was applied. The methodology adopted the Rational Unified Process (RUP), Convolutional neural network and UML tools where applied for the system design. The architectural design consists of three input parameters stored in the hidden neurons, and one output. A back-propagation Convolutional neural network was used to train the parameters. The system was implemented using Hypertext Pre-processor (PHP) programming language. An input interactive interface was generated for predicting parameters threshold values for pipeline intrusion threat ranging from (0-18) pound by square inch(Psi) for threat while (19 and above Psi) for normal. Comparison has been carried out on the outcome between existing system and the proposed system. Results shown in the graph, denoting manual digging, pipeline leakage, walking on pipeline, and pressure. The intrusion point is indicated at line six in the result table where the pressure drops as a result of manual digging. The use of Convolutional neural network in pipeline surveillance system has shown that oil and gas pipeline intrusion can be monitored and controlled.

Vandalism, Prediction, Deep learning, Convolutional Neural Network, Pipeline, Surveillance

[1]. Schmidhuber, J., "Deep learning in neural networks: An overview." Neural networks, Vol.61, Issue.89, pp.85-117, 2015.
[2]. Krzysztof, J. C., “Deep neural networks – A brief historyCao Y, Chen Y and Khosla D. 2014. Spiking deep convolutional neural networks for energy -efficient object recognition”, Intern. Journal of Computer Vision. Vol.21, Issue.51, pp.100-350, 2017.
[3]. C. Adrian, Carlos S., Alejandro R. R., Pascual C., "A review of deep learning methods and applications for unmanned aerial vehicles.", Journal of Sensors, 2017.
[4]. O. G. Chinwe, E. N. Osegi., "An Integrative Systems Model for Oil and Gas Pipeline Data Prediction and Monitoring Using a Machine Intelligence and Sequence Learning Neural Technique.", Vol.6, pp.1-16, 2018.
[5]. Dehghan, Z. (2017).Multi-instance deep learning: Discover discriminative local anatomies for bodypart recognition, IEEE Trans. Med. Imag., 3(5),1332-1343.
[6]. Z. Yan, Y. Zhan, Z. Peng, S. Liao, Y. Shinagawa, S. Zhang, D. N. Metaxas, X. S. Zhou, "Multi-instance deep learning: Discover discriminative local anatomies for body part recognition." IEEE transactions on medical imaging, Vol.35, Issue.5, pp.1332-1343, 2016.
[7]. Yakubu AjijiMakeri, J. Technological Innovation in Crime Prevention in Nigeria. International Journal of Scientific Research in computer science and Engineering; sensor Vol.6, issue .6,pp 66-72.2018
[8]. TejashreePhatak, S.D. Sawarkar;Detection of Faulty Sensor Node within Wireless Sensor Network for improving Network Performance; International Journal of Scientific Research in Network Security and Communication; sensors Vol.5,issue.3.2017
[9]. EIA. Nigeria Country Outlook, Washington, EIA Publications. 2015.
[10]. K. Eero, K. Lukka, A. Siitonen., "The constructive approach in management accounting research." Journal of management accounting research Vol.5, Issue.1 pp.243-264, 1993.
[11]. F. Debo, C. Echem, A. Okoli, M. Mondanos, A. Bain, P. Carbonneau, A. Martey., "A Practical Application of Pipeline Surveillance and Intrusion Monitoring System in the Niger Delta: The Umugini Case Study." In SPE Nigeria Annual International Conference and Exhibition. Society of Petroleum Engineers, 2017.
[12]. Korndoerfer, T. L., "Sustainable Development: A case study of the natural resource use of Yelwa Village, Nigeria." Vol.39, Issue.45, 2009.