All South Africa’s sugar mills are now using the near-infrared spectroscopy (NIRS) analytical technology developed by centralised scientific organisation Sugar Milling Research Institute (SMRI), says SMRI research and development manager Steve Davis.

The SMRI-NIRS technology – which SMRI affiliate member mills in eSwatini, formerly Swaziland, Mozambique, Malawi and Zambia also plan to use from next season – is a part of the current appetite for data-driven Industry 4.0 solutions to improve efficiency and productivity in South African sugar mills.

South Africa has 14 sugar mills, with production of between two-million and 2.5-million tonnes of raw sugar a year.

These mills started using the technology during this year’s sugar season, and the potential value derived from the SMRI-NIRS technology is only just beginning to be explored, SMRI strategic research group leader Dr Katherine Foxon tells Engineering News.

The SMRI-NIRS technology is “extremely quick and easy”, allowing for large numbers of samples to be analysed in a short period, she says, noting that this has huge implications for improved factory control and automation.

“Currently, the conventional analysis of factory streams is time-consuming and expensive with some sample types requiring the use of environment-unfriendly chemicals. Consequently, there has been a trend to reduce sampling regimes with a consequent loss of process information,” says SMRI quality and method development head Stephen Walford.

The SMRI-NIRS technology can analyse many samples quickly and cheaply, allowing greater insight into factory dynamics that did not previously exist. Some analytes that are easily predicted by the SMRI-NIRS technology, such as sucrose and the break-down products of sucrose – fructose and glucose – and sugar stream colour, were not historically measured in most factory streams because of the cost and effort required.

Mills are generating profiles for these analytes across the process, providing new insights into factory operations.

For example, a mill presented an independent paper at the South African Sugar Technologists’ Association annual congress in August last year, describing how the SMRI-NIRS technology had enabled the facility to troubleshoot its centrifuge operation, reducing sucrose losses to final molasses through rapid identification of problems in the centrifuge station, Foxon highlights.

To allow for troubleshooting and optimisation of the technology in mills, the SMRI research team is developing decision-support toolkits and strategies.

The Case for Industry 4.0

Local sugar millers are keen to cash in on the purported benefits of Industry 4.0; however, there are few people with sufficient knowledge and experience to explain how to implement concepts and derive value at factory level, especially in a complex agroprocessing context, says Foxon.

“The sugar industry is dealing with extremely tough market conditions and ageing infrastructure, while the sugar technology knowledge base is also eroding as the mobile workforce makes career sugar technologists a rare and dying breed.”

Industry 4.0 principles aim to address these challenges by enhancing process performance and equipment life by allowing for agile decision-making and the capturing of expert knowledge in models and data systems.

“We believe that these techniques will be essential to enable factories to get the best out of their assets and people while reducing processing costs,” states Foxon.

To drive the competitiveness of the industry, the Department of Science and Innovation, through its Sector Innovation Fund programme, and the country’s sugar millers are co-funding Sugar Factory 4.0 – an about R20-million programme focused on developing Industry 4.0 principles.

Through the funding supplied for Sugar Factory 4.0, SMRI aims to coordinate existing knowledge on sugar technology, data science and industrial Internet of Things by bringing researchers, industrial data science application experts and sugar mills together.

“The research aims to target well-known sugar factory problems that have been unmanageable using traditional control and optimisation methods. These problem areas are complex, have many variables that fluctuate independently and are encountered by most or all mills,” says Foxon.

Sugar Factory 4.0 will allow for solutions to be developed and demonstrated in host mills that can be rolled out to other operations, she notes.

Several of these solutions will take advantage of the increased data availability (analytes and frequency) made possible by the SMRI-NRS technology.

Through the programme, between three and six projects will be undertaken at up to three host milling operations, where a technical solution that is likely to improve efficiency or cost management will be piloted.

“The solutions will not be directly transferable to other mills at the end of the programme, but the knowledge gained from these pilots is expected to reduce the payback period for subsequent projects and reduce the time required for implementation.”

The sugar industry is keen to see Industry 4.0 at work, but is aware that, without proper thought and integration, projects could fail to deliver, she notes.

The SMRI believes that the successful demonstration of Industry 4.0 solutions will create a market for similar projects in the future, as well as a cohort of data scientists with sugar technology knowledge and sugar technologists with data science knowledge, which will improve the ease of undertaking new and more complex projects.

“Job losses are the elephant in the room when people start talking about Industry 4.0. In the sugar industry, the focus is not job losses, but job preservation. We hope that these projects will assist sugar factories to increase profitability at a relatively low cost, as part of a bigger strategy to ensure that the sugar industry can remain a major employer in the KwaZulu-Natal and Mpumalanga sugar cane growing regions,” concludes Foxon.