Adversarial machine learning on neural networks-based control systems
In the area of control systems, the concept of intelligent control refers to the use of AI techniques to control dynamic systems in process engineering. For example, industrial control systems, robots, unmanned vehicles, among others. One of these techniques is Artificial Neural Networks (ANNs). Traditional applications of ANNs (e.g., identification of objects in images) have been successfully exploited through Adversarial Machine Learning techniques. However, it is not clear if these techniques can be applied against ANNs in control systems and what the consequences of such attacks would be. The latter would be the research questions to be addressed in this project.