More and more manufacturers are adopting a new approach to process optimisation by integrating practical advances in technology into their day-to-day operations. The concept is built around developing a plant-wide collaborative process that results in efficiency gains that would otherwise have been missed. Many global tech and industry experts believe that this approach is the way forward in realising the fourth industrial revolution. The concept is called Smart Manufacturing.
This paper, which was written by Opti-Num consultants (Richard Fisher and Jason Miskin) in collaboration with John Thompson (a division of ACTOM), was presented at the IFAC MMM (Mining, Minerals and Metal Processing) 2019 conference. It discusses a practical end-to-end implementation of a Smart Manufacturing technique, namely Model-Predictive Control (MPC) in search for efficiency gains. An unorthodox methodology of MPC development was taken which highlights some of the powerful capabilities of MATLAB, with the benefit of lending itself to usage in other areas such as Predictive Maintenance. The problem was solved in a way that ensures scalability (to other equipment) as well as extendibility (to other techniques).
Jason is a Simulink Consultant at Opti-Num Solutions. He specialises in model-based design and systems and algorithm modelling using the MathWork tools. Jason aspires to promote the adoption of a model-based design workflow in South Africa with the hope that local companies can compete on a global scale.
He completed his Masters in Chemical Engineering at Stellenbosch University which focused on the modelling of a chemical process and associated abnormal events, prior to working at Opti-Num Solutions.
Richard is a Consultant at Opti-Num Solutions, specialising in algorithm and workflow development, with special interests in machine learning and mathematical modelling. Richard holds a BSc(Eng) in Mechanical Engineering from the University of the Witwatersrand. He has recently completed his MSc in Computer Science in the field of robotic path planning in dynamic environments, which incorporates machine learning, mathematical modelling and problem solving.
Richard aspires to use data analytics and sophisticated algorithms to solve real-world problems in creative and novel ways.