Coastal nations in Africa increasingly have to deal with numerous threats at sea such as piracy, illegal fishing practices and the pollution of marine life. And they don't always have the capacity and ability to detect these dangers early and to guard their maritime jurisdiction areas.
"Coastal nations are hamstrung by a lack of threat detection and threat evaluation intelligence, a very large jurisdiction area, a limited amount of maritime law enforcement resources, a vast quantity and variety of potentially threatening activities, and conflicting goals and objectives on how maritime law enforcement resources ought to be allocated," says Dr Alexandre Colmant who recently obtained his doctorate in Industrial Engineering at Stellenbosch University (SU). His supervisor was Prof Jan van Vuuren of SU's Department of Industrial Engineering. Colmant currently works as a data analyst in Cape Town.
As part of his doctorate, Colmant developed a semi-automated decision support system that could contribute towards the efficient allocation of maritime law enforcement resources such as patrol and military vessels and armed helicopters to investigate potential threats in real-time. This could also help prevent many people and/or organisations getting away with committing illegal activities at sea.
"Because the number of potential maritime threats can be overwhelming and the nature of the decision process is typically complex, the quality of resource assignment decisions can be improved significantly by providing maritime operators with computerised support."
Using mathematical modelling principles and four higher-level techniques, Colmant tested the decision support system in two realistic, but simulated maritime law enforcement scenarios of varying complexity under different computational times.
"It was found that, based on the complexity of the scenarios and the amount of computational time, certain search techniques are better than others at finding a high quality set of alternatives or filtered set of all feasible solutions that human operators can use during the complex decision-making process."
Colmant adds that effectively and promptly filtering through the many alternatives at any given time to ultimately identify the one most preferred by operators can only be achieved with the help of computerised support.
"My decision support system is expected to help maritime law enforcement agencies conduct operations at sea more effectively which will lead to a higher rate of threat neutralisation, a lower rate of uninvestigated detected potentially threatening activities, lower operating costs and a higher level of consensus amongst the various maritime decision-making bodies."
"Operators may use the system as a guideline to validate and/or justify their decisions, especially when the level of uncertainty pertaining to the observed maritime scenario is high."
Colmant says the use of the system in a real-world context may also reduce operators' stress levels typically associated with making difficult decisions.
He points out that although the system could in real-time determine much easier than human operators what the probability is that a given object or activity at sea embodies a specific type of threat, this would not be due to the incompetence of human operators.
"Although there may be less reliance on humans to make important decisions regarding potential threats at sea, it does not entail that such human operators will become redundant and out of a job."
"In fact, the combined expertise of these operators to be used as various forms of input in the decision support system is imperative."
Colmant says research has shown that the decision-making process involved in maritime law enforcement response selection by developing coastal nations is often exclusively conducted by humans without the use of automated decision support systems.
"These decisions are usually made by one or more individuals, based on intuition and experience in the maritime law enforcement domain in contrast to being the result of an automated, analytical process."
According to Colmant, human operators may use the decision support system prior to running the search algorithms to identify and limit the range of options for decision making at their disposal.
"This will save significant computational time as well as increase the likelihood of finding good alternatives."
Colmant says the decision support system still needs fine-tuning and more research has be done in terms of demonstrating its strengths to the interested entities as well as possibly uncover certain ways to improve or modify it.
"The system still has to be tested against a group of human operators who solely rely on their experience in the field."
Regarding possible future use of his system, Colmant says he has been in close contact with the Institute of Maritime Technology in Simon's Town, and also presented his findings to ARMSCOR in Pretoria.
Colmant adds that he would still like to continue his research on a part-time basis and publish his findings in academic journals.
- Photo 1: Credit: Pixabay
- Photo 2: Dr Alexandre Colmant (left) with his supervisor Prof Jan van Vuuren at SU's 2016 December graduation.
FOR MEDIA ENQUIRIES ONLY
Dr Alexandre Colmant
Tel: 021 8764348