??Prof Tobi Louw from the Department of Chemical Engineering in the Faculty of Engineering at Stellenbosch 肆客足球 delivered his inaugural lecture on Tuesday 19 August 2025. The title of his lecture was “A critique on the application of machine learning in chemical engineering".
Louw spoke to the Corporate Communication and Marketing Division about his research on the application of machine learning in the chemical and minerals processing industries.
Tell us more about your research and why you became interested in this specific field.
I'm interested in understanding new developments in machine learning and investigating ways in which they can be used to control and optimise processes in the chemicals and minerals industry. I've always enjoyed applied mathematics, but things really clicked for me during my undergraduate studies when I first encountered process modelling and control. Seeing how mathematics can be used to analyse the dynamics of complicated chemical processes I've been studying was wonderful. I've been fortunate to have the opportunity to continue working in the field as an academic.
How would you describe the relevance of your work?
Chemical engineers design and maintain the industrial processes that convert raw materials to valuable products, but these processes don't run smoothly all the time. Process monitoring and control are necessary to ensure the processes operate safely and efficiently. These plants generate large volumes of data, which can be analysed using modern approaches to further optimise the process: that is where my work comes in.
Can you give examples of how your research is applied in real-world contexts?
Process engineers often find that the process isn't operating quite as it should — this may indicate the presence of a fault. It's usually difficult to find exactly where the fault is coming from: does the problem lie with a worn-out valve, or fouling on a heat exchanger? Machine learning can help with this: by analysing patterns in the data, the faults can be identified and the problem addressed.
What makes it difficult to implement and sustain machine learning technologies in the chemical and minerals processing industries?
I think there are two major challenges. First, it's often quite difficult to get good measurements under very harsh process conditions: high temperatures, corrosive gases, etc. Without good measurements, analysing the data becomes much more challenging. Second, machine learning tools have become easily accessible in recent years, but traditional engineering training does not equip engineers with the statistical knowledge to correctly interpret the results. Providing industry practitioners with the necessary know-how is essential.
Looking into your crystal ball, what developments do you see in machine learning?
Large Language Models (LLMs) have really changed the way we think about machine learning. There are many ways in which LLMs can impact the processing industry, not the least of which is coding support. LLMs make it easier than ever for process engineers to start coding and developing their own data analytics tools. However, there are also really interesting developments in the use of neural networks for time series forecasting — that is, to predict process conditions. This will again be limited by available measurements. I think the use of machine learning to support the development of new “soft sensors" is very important for the field. “Soft sensors" use easily measured values to infer the properties of a system: for example, video imaging may be used to analyse how well a crusher (a machine that crushes big chunks of material into smaller ones) is performing.
Being part of a Machine Learning research group, what aspects of your work do you enjoy most?
I really enjoy studying the statistics and mathematics that underpin the various machine learning methods. But the best part of my work is definitely engaging with students and seeing how they develop in their ability to critically evaluate these complicated algorithms.
The higher education environment can be challenging. What keeps you motivated when things get tough?
As a Christian, I believe we have a sacred task of “ruling and subduing the earth", which in modern parlance translates to wisely stewarding our resources. Chemical engineers create great value out of seemingly valueless raw materials: think of platinum from rocks. I think this is a worthy vocation, and we have a duty to do so in a way that is socially and environmentally responsible.
Secondly, I believe all people are created in the image of God. Knowing that the people you work with are wonderfully created and deeply loved has a real impact on my daily interactions, even though I still get it wrong much of the time!
These are the two cornerstones that keep me motivated: recognising the intrinsic value of the work that I do and the immeasurable value of the people that I work with.
Tell us something exciting about yourself that people would not expect.
I help organise a film club that meets monthly to watch a movie and discuss the ways in which the movie reflects our culture and our world; sometimes I even facilitate the conversation afterwards. We watch a wide variety of movies, from Barbie to Dunkirk, but my favourite by far is the 1987 classic Babette's Feast.
How do you spend your free time?
I love playing board games. We have a board game group that meets every Sunday evening. We are a very competitive group, but also really good friends. I'm training my kids to be board gamers too.
?
