The South African Council for Automation and Control will be hosting a workshop showcasing industrial applications of machine learning in process monitoring and control (details below) as well as providing hands-on experience in the application of basic ML techniques. Participants will be guided through examples illustrating fundamental concepts of ML, including the bias-variance trade-off, linear regression, classification, resampling techniques, regularization and feature extraction. The examples are designed to clearly illustrate ideas that are relevant to all ML applications, providing exposure and resources to enable participants to engage with the field more effectively. The workshop will culminate in participants applying what they have learnt to develop a soft sensor for a well-studied penicillin fermentation reactor. The workshop is sponsored by Stellenbosch University’s Faculty of Engineering
The workshop is aimed at newcomers to the field of ML, but some basic programming knowledge is required.
Workshop attendance is free for conference attendees
More information on this event
ROELOF COETZER obtained his PhD in Mathematical Statistics from the University of the Witwatersrand in 2004. With 30 years’ worth of industry experience (including a term as President of the South African Statistical Association). He has been successful in driving Data Science projects and Big Data Analytics solutions, leading highly technical multidisciplinary teams and developing technology packages. He has co-authored 47 peer-reviewed articles in national and international scientific journals and conference proceedings. Prof Coetzer joined the North-West University as Associate Professor in the Department of Statistics in June 2021. His industrial case study is titled Specifying process health indices for multivariate process monitoring and diagnostics using machine learning models.
JACQUES STRYDOM graduated with distinction from Stellenbosch University with a B.Sc. Hon in Computer Science. In 1998 he started his extended career at Sasol and played a key role in the progression of an Advanced Process Control and Optimisation footprint. He was promoted to Chief Engineer: Control and Instrumentation in 2012 and soon after joined Sasol Group Technology’s Centre of Expertise as Principal Specialist: Process Control and Optimisation. He is currently appointed as Principal Specialist: Digitalisation in the Secunda Operations division of Sasol’s Energy Operations. His industrial case study is titled Continuous control of an industrial separation process using an empirically derived virtual analyser.
OPTI-NUM SOLUTIONS collaborates with a wide range of industries to identify problems and find workable solutions that empower their clients to make data-driven, evidence-based decisions. Opti-Num’s clients benefit from their multi-sector expertise augmented with the modular MathWorks tools. Smart Mining & Manufacturing, a division of Opti-Num Solutions, aims to create value across the supply chain through their extensive knowledge of data science, advanced data visualization, smart monitoring and maintenance, and intelligent control. In this talk, titled Soft-sensing boiler health through machine learning, Opti-Num will demonstrate how an intelligent soft sensor was developed using ML in the sugar industry. This information is used by operators to take appropriate actions to prevent possible trips. The soft sensor could possibly be used within Model Predictive Control in future to automate intervention.
STONE THREE an Industrial Internet-of-Things company that develops AI-augmented solutions within the digital productivity, workplace safety and employee healthcare sectors. They offer smart sensors that leverage machine vision technology, as well as offering process monitoring services: generating actionable advisories for improved process performance. In this talk, titled Industrial machine vision with deep learning: Challenges and best practices Stone Three will share their insights on the challenges and best practices for scalable machine vision smart sensors: including what is required for industrial image data collection, image labelling, label review, model training and review, model deployment and model updates.