Manufacturing innovation driven by the intelligence of network-based information technology also known as smart manufacturing is a key to achieving sustainability, triggering unprecedented economic growth, and comprehensively improving competitiveness. The fundamental elements required to achieve smart manufacturing are (a) model-based, data-centric foundations (b) real-time optimization algorithms (c) model-based control strategies and (d) high-performance, low-cost microcontroller platform. Smart manufacturing relies heavily on the physical infrastructure to collect quality relevant data and analyze information across multiple scales. Integration of novel data analytics, predictive modelling capabilities and optimal control strategies allows radically better ways of operating chemical process plants. Modelling provides an unprecedented operational insights while control automation helps achieve environmental sustainability, operational performance and economic profit margins. Therefore, the main challenge of our research program is to test the smart manufacturing capabilities by performing rigorous experimentation on the laboratory scale hardware platform which replicates real world operating conditions. Towards that goal, we have built a fully automated and highly instrumented experimental test platform to study the multi-scale fuel cell phenomena. We believe that the intelligent decision making underpinned by the smart manufacturing applied to the fuel cell power plant can demonstrate the absolute limits of performance improvements. The integration of these elements thus offers a step change in operational excellence against the incremental progression achieved on their individual merits.
Fuel cell involves a dynamic phenomena comprising of complex interactions of mass transfer, energy transport and electrochemical kinetics. It is a demand response technology that presents big data challenges, when resolved can pave the way to a greater operational insight across its value chain. In that regard, our research focuses on two pronged agenda (1) carry out fundamental experimental understating of fuel cell science and (2) experimentally validate existing and new process system engineering tools on a platform that is representative of industrial environment. Thus, the ultimate goal of our research effort is to provide a system level perspective under a real industrial setting to collect and analyse data reliably across multiple scales from unit-to-stack-to-plant that can be processed by intelligent decision making optimization and control algorithms empowered by multi-parametric programming. In this presentation, we describe a unified framework for the smart manufacturingthat achieves seamless integration of data, process and technologies for the fuel cell energy system. Our framework uses (a) high fidelity mathematical modelling at multiple scales to capture complex process dynamics, (b) optimization tools and techniques to analyze impact of uncertainty, (c) model predictive control strategy to deliver on performance, safety and economics. In the presentation, we highlights the results achieved using a novel in-situ experimental technique to understand the inner workings of a fuel cell at unit cell level. This understanding is directly linked to the dynamic modelling and validation of the fuel cell stack and the integrated balance-of-plant. Finally, we present a step by step procedure to deploy “MPC-on-a-Chip” that delivers the outlined benefits of model-based innovation for the smart manufacturing of hydrogen energy.
 The financial support from EPSRC grants (EP/I014640/1 and EP/K503381/1), sponsorships from CPSE industrial consortium and Texas A&M Energy Institute is gratefully acknowledged.
 Note that this work is generic and can be easily extended applied to any type of fuel cell or chemical processing plants.
See more of this Group/Topical: Topical 3: Innovative Manufacturing