Simon M. Miller, De Montfort University
Supply Chain Modelling: Representing Uncertainty
Planning resources for a supply chain are a major factor determining its success
or failure. Poorly managed resources can result in not being able to satisfy customer orders
(stock outs) and holding stock unnecessarily.
In his research, Simon used fuzzy logic to model the supply chain allowing him to
represent uncertainty while making sensible resource planning decisions. Instead of
using crisp values, fuzzy logic allows the description of data using intuitive linguistic
categories. Using a fuzzy model, can achieve more robust resource plans, as it does
not insist on crisp input values.
The model has been used to look for a good resource plan for an example
problem. To find a solution a genetic algorithm has been used. The algorithm
searches for a resource plan that satisfies a supplied forecast, with minimum cost,
at a given customer service level.
The proposed model can account for the fact that expectations may be
uncertain, vague or incorrect. Using a fuzzy model in this way can ensure that
plans are still appropriate, even when reality differs from predictions. The benefits
to the user of such a model are a reduced need for safety stock, as they can have more
confidence in their plans, and reduced stock outs minimising lost sales and customers.
Darren Ghent, University of Leicester
Is Africa a source or sink of carbon?
Global warming driven by increases in greenhouse gas emissions is currently a hot topic in the media. With this
attention fuelling growing public concern regarding the impacts of climate change, it is important to ensure uncertainties
in the science are minimised in order to accurately quantify the effects, and to inspire confi dence in the predictions of
future change. Much emphasis has recently been placed on quantifying human caused emissions of greenhouse gases, such
as burning fossil fuels and deforestation. Less attention has, however, been given to the natural fl ux of ecosystem carbon,
termed net ecosystem productivity (NEP), and how this changes from one year to the next.
Africa is the least studied continent, and is likely to experience the greatest impacts from climate change. Unfortunately,
climate scenarios for the continent are highly uncertain. One of the main drivers of climate variability in Africa is El Niño,
whereby a strong event causes drier than average conditions during the wet season. It is important to identify these changes
before accurate predictions can be made of the long-term trends due to increased atmospheric carbon dioxide, and whether
the land is likely to act as a source or sink of carbon. By applying a climate model, this research indicates Africa behaving as a
source of ecosystem carbon in years experiencing strong El Niño events.
Although fossil fuel emissions for Africa are relatively low on a global perspective, as the world warms, many models
suggest more frequent and intense El Niño events will result, suggesting more emphasis should be placed on African
emission sources.
Elizabeth Massey, University of Lincoln
Automated detection and diagnosis of diabetic retinopathy
Diabetes mellitus is a chronic disease affecting over 150 million people worldwide and around 2.35 million
people in the UK alone. Long-term complications can occur to the different systems in the body, including
the heart, kidneys, blood vessels and the eyes. The ocular structures most affected are the lens and the
retina.
In the retina, diabetes can lead to an eye disease known as diabetic retinopathy, which has been identifi ed
as one of the leading causes of blindness in the working population of industrialised countries. A person with
diabetic retinopathy may have serious retinal damage before ever noticing visual loss. Therefore, it is important
for diabetics to undergo regular screening by an ophthalmologist.
Diabetic retinopathy can be prevented with a good diet and early detection. During annual screening, clinicians
and doctors spend many hours pouring over retinal images looking for indicators of DR; lesions that are
sometimes sub-pixel size. In Lincolnshire alone, over 36,000 patients are screened annually. That’s over 144,000
patient images to be analysed. That is a daunting task for even the most skilled clinician.
This is where Elizabeth’s work comes into play. She is looking at ways to build software algorithms that
will help doctors and clinicians with their analyses. As part of her PhD research programme, Elizabeth
is building math models and working with neural networks to not only detect the existence of the
disease but also specify the type and severity. Overall, the system endeavours to cut the cost of
screening and provide patient information in a timelier manner.
Engineering Designer July/August 2009
25
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36