Preparing Industry for IIoT: Separate Sensor Networks for Industrial Automation Security

Abstract

Industrial automation has brought innovations in efficiency and safety to modern manufacturing. The emergence of big industrial data, AI and Internet-connected industrial automation systems promises many advantages in manufacturing, maintenance, health & safety, product customization, logistics, and reporting. The Industrial Internet of Things (IIoT) seeks to attach industrial systems to the Internet for data collection and processing. The Internet is not a secure place for many of the industrial systems in use today. Industrial networks are designed for speed, accuracy, and availability at the expense of security. In our experiments, we demonstrate how easy it would be for hostile actors to gather data that could be used to compromise the industrial automation systems if actors (a) had physical access to the internal network or (b) access to data streams that feed to the external network. To resolve some of these issues, we propose a hybrid solution of sensor data acquisition in which a secondary network with additional sensors and controllers are installed; the secondary network, sensors and controllers are isolated from the internal control system. The results show that our solution enables useful data collection without impacting the security, speed, accuracy, and availability of industrial control system. It also has the potential to permit anomaly detection using independent sensors on different networks.