Industrial Safety Management using Control Theory and Machine Learning

Optimizing plant processes is of prime importance now more than ever. With stricter infrastructures being placed on safety, environmental effect, and corporate social responsibility, more complex systems that optimize these factors are needed. These complex systems with advanced algorithms are intended to further streamline the existing process while mitigating issues leading to a safer workplace, and environment, while creating cost savings potentially in the millions. Two of the biggest problems that are hindering this growth are process optimization and management of alarm systems. This proposal aims to incorporate a methodology that combines traditional safety analysis techniques with recent data-driven techniques such as machine learning and reinforcement learning to minimize and manage the alarm system and suggest operational set points to increase throughput while optimizing safety. The proposed research will use data from the plant process at NTWIST Inc will be used to obtain optimal policies that minimize cost associated with the unsafe operation.

Faculty Supervisor:

Mo Chen

Student:

Sriraj Meenavilli

Partner:

NTwist

Discipline:

Computer science

Sector:

University:

Simon Fraser University

Program: