Niket S Kaisare

Advanced Process Control: From Industry Inspiration to Industry Implementation

Between two tenures as a faculty in IIT Madras (2007-2011 and 2014-now), I spent three-and-half years in industrial R&D in Bangalore. In this talk, I intend to take the audience on a journey into a research that was inspired by and another that was implemented in the industry. While mathematical details are intriguing and practical implementation is exciting, I take a different approach in this talk; this talk sits at the boundary of intriguing and exciting!

The first example involves a “human factors study” to understand how humans interact with an automation system. Today’s chemical plants are highly integrated and are controlled with advanced techniques such as model predictive control (MPC). Whether it is self-driving cars, nuclear reactors or process plants, human factors studies often focus on abnormal situations. However, human drivers or operators spend significantly longer time in normal situations. Here, we analyze human behavior when interacting with advanced control systems. Specifically, we allow operators to interact with a control system through a human-machine interface, measure their performance in completing a task in controlling a process plant, and monitor their eye-gaze and other behaviors.

The second example comes from a field study in shale gas automation. Here, I discuss our experience in commissioning a control system in a shale field in the deep winter of Alberta, Canada. The challenges in running an automation algorithm trained using data from Texas field (typically at 35 deg C) in frigid Alberta winter (minus 50 deg C) will be discussed.