Senior Design Team sdmay22-38 • ADSICS Anomaly Detection System for Industrial Control Systems

Summary:

ADSICS is a surveillance program that enables power distribution companies to detect, analyze, and prevent network anomalies that could threaten their security. It processes alerts and displays them to the companies’ cybersecurity professionals in an intuitive manner. To complete this project, we have researched about cybersecurity basics, researched and worked with the power grid testbed, performed data augmentation using existing open-source dataset and actual data, and experimented with machine learning algorithms on the testbed environment. We aim to use a machine learning algorithm within our VM to process alerts, ensuring that they enter the alert correlation algorithm with a uniform format and contain all the necessary information (time, location, type) to be analyzed by a user.

Introduction to our platform and Elastic:

Demo of Elastic and ADS:

Demo of Elastics Machine Learning Data Frame Analyis:

Demo of Elastics Security Component and Our Dashboard: