Over the past ten years, the Institute of Systems Science (ISS) at the Durban University of Technology (DUT) has been quietly contributing to a range of scientific fields through its rigorous research initiatives where members specialize in modelling and simulation.
The ISS has several associated scientists, some with B and C ratings from the National Research Foundation (NRF). These people have contributed significantly to our efforts as a team and have published several publications on a broad range of projects in applying models to practical questions from the mathematical, applied, natural, social, and agricultural sciences. I head the institute and have a South African Research Chair (SARChI) funded by the NRF.
The research of the institute focuses on diverse topics such as the impact of resource limitation on habitat usage by different species in Hluhluwe-iMfolozi Park, the influence of animal movement on population dynamics, and the dynamics of city populations under different conditions, and even studies on the workings of the universe.
Investigations in wildlife ecology examine how large herbivores, such as African elephants, adapt their movement patterns in response to environmental changes. This research is crucial for developing targeted conservation strategies that accommodate the natural behaviours and needs of wildlife, particularly in regions affected by climate variability and human encroachment. In one example, it is shown that directional movements that are predominantly forward are more beneficial to animals in patchy resource environments, potentially enhancing survival during environmental perturbations. Other studies explore how resource depletion and competition affect habitat usage among a guild of browsers like elephants and giraffes in a South African park.
Moving to urban studies, the institute has explored the socioeconomic drivers behind urban migration, with specific attention to how income levels influence movements to and within major South African cities. This research provides valuable insights for urban planners and policymakers, helping them to manage urban growth and demographic changes more effectively.
Some of the ISS team have conducted a series of studies to improve on an understanding and management of various infectious diseases affecting both agriculture and wildlife. The research focused on the dynamics and control mechanisms of diseases such as Bluetongue virus and Maize streak disease, as well as examining the complex interactions at wildlife-livestock interfaces, such as those involving African Swine Fever.
In each study, different mathematical and epidemiological models are developed to assess the effectiveness of various control strategies, including vaccination, quarantine, and insecticide use. The goal is to identify the most cost-effective and practical methods to curb disease spread in affected regions, contributing to global efforts in disease management and prevention.
In recent years the diseases studied are important to sustainable agriculture. Examples, include the following diseases. Bluetongue disease is a viral infection affecting ruminants like sheep and cattle, characterised by symptoms such as fever and a blue-tinted tongue; African swine fever is a highly contagious disease impacting pigs and wild boars, with severe consequences for the swine industry. Lastly, maize streak disease is a viral infection that targets maize crops, marked by yellow streaks on the leaves and stunted growth, leading to substantial crop losses.
The research aimed at enhancing understanding and managing these outbreaks more effectively. For Bluetongue disease, mathematical models were developed to analyse the transmission dynamics and evaluate the effectiveness of various control strategies, such as vaccination and quarantine measures, aiding in minimising its spread across different regions. In addressing African Swine Fever, the work focused on assessing the transmission cycles involving wildlife and livestock interfaces. By identifying critical control points and the role of vectors such as ticks, these studies provide insights that assist in formulating targeted interventions to prevent and control outbreaks. Regarding Maize streak disease, stochastic models were developed to explore the dynamics and control options of the disease. This research points to strategies that combine mechanical and chemical measures to effectively reduce the incidence of the disease, thus minimising the impact on maize yields and ensuring food security in affected regions.
Other research is on non-communicable diseases. For example, in the field of dermatology, where the group has applied artificial intelligence to improve the accuracy of skin disease diagnoses. The aim is to diagnose skin cancer from skin lesion images using computer-aided diagnosis (CAD) systems, which are crucial for early detection and treatment. The team highlight the effectiveness of convolutional neural networks (CNNs) in medical image segmentation and classifications. They demonstrate that their optimised CNN model outperformed existing methods on multiple metrics, making it a promising tool for clinical application. In related work, research has been completed on improving dental image segmentation using a local ternary pattern encoder-decoder neural network. This study focuses on enhancing the accuracy and efficiency of segmenting dental images, which is crucial for the effective diagnosis and treatment planning in dental care.
In research that overlaps environmental issues with diseases the ISS, in collaboration with Wageningen University in the Netherlands, has conducted research on deformed wing virus that can severely affect honeybees. Transmitted mainly by Varroa mites, it distorts bees’ wings, often preventing flight, and reduces their lifespan, threatening colony health and the critical pollination that bees provide. A stochastic model is developed to investigate how grooming and hygienic behaviours in honeybees can mitigate the impact of deformed wing virus. By applying branching process theory, the model demonstrates that these behaviours significantly reduce the likelihood of virus outbreaks, highlighting the potential of selective breeding for these behavioural traits to enhance colony resilience.
Finally, in terms of the environment, the institute has addressed the environmental impacts of industrial activities, specifically within the cement production sector. By conducting comprehensive reviews of life cycle assessments provides a broader understanding of how industrial practices can be modified to reduce environmental footprints and promote sustainability.
In applications to engineering systems research on path loss prediction for 5G and beyond focuses on developing and comparing advanced machine learning models that can accurately predict signal loss in high-frequency bands, specifically in challenging indoor environments. The success of these models is measured by their ability to closely fit the actual measurement data, which is crucial for optimising future wireless networks to ensure reliable and high-quality service.
In other work a detailed symmetry analysis of a 2D traffic-flow model is used to explore traffic dynamics on two-lane motorways and a Lie symmetry analysis is conducted on a model for shallow-water systems, focusing on sediment dynamics.
Other studies investigate the evolution of cosmological anisotropies within a model cosmology using scalar-torsion theory and explore bulk viscous scenarios in cosmology using data from Cosmic Microwave Background and Type Ia supernovae, enhanced by simulations of gravitational waves. Additional research examines anisotropic solutions in model universes within modified teleparallel gravity, discussing their instability and theoretical implications.
Other members of the group study phenomena in classical and quantum physics, including the theory of open quantum systems, quantum optics, theory of gravity, the physical vacuum, cosmology, astrophysics, physics of metals, geophysics, atmospheric and plasma physics. For these studies non-linear and non-Hermitian Hamiltonian approaches are used.
Through these diverse research efforts, ISS continues to impact a wide array of scientific disciplines, providing critical results that can potentially lead to innovative solutions for some of today’s pressing challenges.
Pictured: Professor Kevin Duffy
(Professor Kevin Duffy currently works at the Institute of Systems Science, Durban University of Technology. Kevin does research in Ecology, Agronomy and Applied Mathematics).
