Process Systems Engineering
at Louisiana State University

Project: A Federated Sensor Network Architecture for Distributed Data Analysis

Students: Y.Y. Joe (Singapore), B. Bhushan (Sydney), R. Willis (LSU)
Supervisors: J.A. Romagnoli (LSU), K.V. Ling (Singapore), A. Palazoglu (UCDavis)

The objective of profitable operation in today’s manufacturing plants is intimately tied to the multi-level management of information regarding supply-chains, physical assets and personnel deployment that requires a nimble, robust and efficient plant environment. Such an environment ensures persistent quality of products, and, by properly managing abnormal operating situations, is expected to respond vigorously to potential threats to the environment and the safety of the personnel and the community at large. The paradigm of proactive information management fundamentally relies on the underlying sensor network that collects and communicates the pertinent process data to appropriate levels within the plant and the enterprise hierarchy. The ability of the sensor network to respond to this challenge is being greatly enhanced by recent breakthroughs in microprocessors-based instrumentation and digital communications that enable the sensors with new functionality and smart sensing capabilities.

The proposed research is based on the concept of federated processing, which deals with multiple processors encapsulated in one machine or system. Federated sensor networks are autonomous, distributed and cooperating entities. In other words, they work autonomously towards their individual goals; they are distributed into different parts of a system or machine and they participate in real-time collaboration to reach the common (system) goal. Such an environment provides a radical shift in the way information management has been perceived and practiced within a manufacturing plant. The complex tasks of plant-wide monitoring, fault assignment and prevention of abnormal events become a realistic, achievable goal.

Back to Project List >>