Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 23

Mathematical Modelling of everyday processes

PROJECT PLACEMENT PROVIDER | KEELE UNIVERSITY: SCHOOL OF


COMPUTING AND MATHS

July 20, 2015 – August 14, 2015


Adams’ Grammar School | Anish Parpani
Abstract
Mathematical modelling techniques are pivotal in understanding the behaviour of a system. The
practical limitations of real world situations can be overcome by forming models. They can be used to
analyse the impact of changes to different variables and can be refined in order to obtain a useful conclusion.
This report looks to investigate two such models. The first requires an algorithmic approach in order to
improve efficiency of a milk delivery system. The consideration of different strategies using Prim’s and
Kruskal’s algorithms allows different delivery systems to be implemented. The second involves the use of
trigonometry and developing strategies to minimise the cost and inconvenience to ensure optimal watering
of a cricket field.

P a g e 1 | 19
Table of Contents
ABSTRACT...................................................................................................................... 1
INTRODUCTION ............................................................................................................... 3
AIMS AND OBJECTIVES: ...................................................................................................... 4
METHODS AND MATERIALS (MILK DELIVERY) ............................................................................ 5
RESULTS FOR MILK DELIVERY ............................................................................................... 8
DISCUSSION OF RESULTS.................................................................................................... 10
METHODS AND MATERIALS (WATERING A FIELD) ....................................................................... 11
RESULTS (WATERING A FIELD) ............................................................................................. 13
DISCUSSION OF RESULTS (WATERING A FIELD) .......................................................................... 15
CONCLUSIONS – MILK DELIVERY ......................................................................................... 16
CONCLUSIONS – WATERING A CRICKET FIELD .......................................................................... 16
CRITICAL EVALUATION AND SCOPE FOR FURTHER RESEARCH ........................................................ 17
BIBLIOGRAPHY ............................................................................................................... 18
ACKNOWLEDGEMENTS ..................................................................................................... 19

P a g e 2 | 19
Introduction
Mathematical modelling of real world situations in order to improve efficiency and provide the
optimum use of resources often requires the use of algorithms. The task assigned consisted of producing
two mathematical models. Firstly, advice was to be provided in order to achieve an efficient delivery of
milk from dairy farms to local homes. The key phrase to note is an ‘efficient delivery’. The technical
definition of this phrase means that it is the process of using the minimum level of inputs to produce the
maximum output. This not only includes physical inputs but also aspects such as time and energy. In
essence, it is the optimal use of the resources available. Within this case study, the definition of being
‘efficient’ can range from aiming to reduce costs to trying to maximising profits.
A firm would look to maximise their profits which might be done through only delivering at off-peak
times (after 8am) allowing lower wages to be paid. However this would have the drawback of customers
deciding to leave the milk delivery company due to the late delivery times. Alternatively the client might
place more emphasis on fuel efficiency based on the road condition and traffic. Moreover they could
become more efficient by reducing the time taken for delivery which might mean buying more expensive
capital equipment, thus increasing their costs and reducing profits. One way this could be achieved is by
switching from electric delivery vehicles to those that use petrol and diesel. An efficient delivery can also
aim to improve the customer experience. This could mean fewer disruptions to their schedule and
delivering the milk on times to suit the consumer. Consequently, there are numerous ways of recognising
the term ‘efficient’ and different views can be considered with different models in order to reach a justified
conclusion. These mathematical models are the key to understanding how to improve efficiency and this is
the basis for this project.
The second model investigated consisted of placing sprinklers in a cricket field to minimise the cost
and inconvenience but to ensure optimal watering. As with the previous model, the wording of the
question is important, especially of the word ‘inconvenience’. Firstly, the players must be considered.
There will be times where the field will need to be watered before the start of the games and during the end
of an innings. This will obviously interfere with players who wish to train and warm-up for their innings.
Furthermore if sprinklers are placed close to the edge of the boundary, there would be an
inconvenience to members of the public who are in the stands. Moreover since the outfield and the actual
wicket will require different amounts of watering, there must be no overlap between the sprinkler which
waters at a high precipitation rate and the wicket as this requires less water. The outfield and the central
wicket will require different amounts of water as “cricket squares are densely compacted, which reduces the
rate at which water can penetrate the soil profile. It is therefore essential that water is applied at a
sufficiently low rate to prevent run-off & wastage of water. Pop-up sprinklers are generally unsuited to
irrigating cricket squares as they apply water too heavily to be absorbed.”1

1
LS Irrigation - Cricket Pitch Irrigation Systems & Services. Available
at: http://www.lsirrigation.co.uk/sports-irrigation/cricket-pitch-watering-irrigation/cricket-pitch-
irrigation.html. [Accessed 03 August 2015].

P a g e 3 | 19
Aims and Objectives:
In this investigation, I have two key aims:

1. Research and investigate the most efficient method for milk delivery to different neighbourhoods. The ideal
approach to this is to start with a very basic model; in this case a simple grid layout. I need to research
specific figures such as wages costs, walking speed, distances and the cost of electricity for milk floats in
order to provide the time taken for delivery in each model as well as the cost. Once basic models have been
developed and analysed, the next step is to apply this to a real world situation. To provide a detailed
analysis, I will look at real world street layouts in isolation as well as together in the same neighbourhood.
This will provide a potential client with a more realistic estimate of time and cost restrictions.
2. Investigate different arrangements of sprinkler systems in order to minimise the cost and inconvenience to
ensure optimal watering of a cricket field. The first task will be to conduct the required research on
sprinkler systems and theoretical approached to packing circles in a bigger circles. I will also need to revert
back to the key principles of trigonometry and apply them in a new situation.

The first step for both tasks is to conduct the necessary research to build and analyse basic models showing these
situations with minimal limitations. As the models develop, mathematical techniques will be used to provide
evidence; justifying the conclusion presented.

P a g e 4 | 19
Methods and Materials (milk delivery)
Definitions:
Greedy algorithms: a mathematical process that looks for simple, easy-to-implement
solutions to complex, multi-step problems by deciding which next step will provide the
most obvious benefit.
Graphs: a graph consists of points (vertices or nodes) which are connected by lines
(edges or arcs).
Path: a path is a finite sequence of edges, such that the end vertex of one edge in the
sequence is the start vertex of the next, and in which no vertex appears more than
once. This can be seen on the graph in Example 1.02 with one possible path being ABDE. Example 1.0
Another possible path could be CEAB.
Cycle: A cycle is a closed path (i.e. the end of the last edge is the start of the first, Vertex
Edge
and no vertices are repeated except that the final vertex is the same as the first). A
cycle that passes through every vertex of a graph is called a Hamiltonian cycle and a
graph in which a Hamiltonian cycle exists is said to be Hamiltonian.3
Tree: a tree is a simple connected graph with no cycles.
Minimum Spanning Tree (MST): the total length of its edges is as small as
Example 1.1
possible. 4

The two main algorithms of use with this problem are Prim’s and Example 1.2
Kruskal’s algorithms. The following steps outline Prim’s algorithm5:
1. Choose any vertex to start the tree.
a. Select an arc of least weight that joins a vertex that is already in the tree to a vertex that is not
yet in the tree.
b. If there is a choice of arcs of equal weights, choose randomly.
2. Repeat step 1 until all the vertices are connected.
Figure 1.0 shows an example of Prim’s Algorithm. As the algorithm states, we can
choose any vertex to start with (vertex d). The least weighted arc is “da” therefore Figure 1.0
this is selected next and so forth. The total MST = 39.
Figure 1.0

2 Decision Maths Tools | Mathematics, Learning and Technology. 2015. [ONLINE] Available
at: https://colleenyoung.wordpress.com/2009/11/06/decision-maths-tools-2/. [Accessed 11 August
2015].
3UCL - D1_graphs_glossary Available at:

http://www.ucl.ac.uk/~ucahbtw/docs/d1lesson1/D1_graphs_glossary.pdf
4 Definition from WhatIs.com. Available at: http://whatis.techtarget.com/definition/greedy-algorithm.

5 Susie Jameson, 2010. Edexcel AS and A Level Modular Mathematics D1. Edexcel Limited.

P a g e 5 | 19
Kruskal’s algorithm has the following steps:
1. Sort all the arcs (edges) into ascending order of weight.
2. Select the arc with the least weight to start the tree.
3. Consider the next arc of least weight.
a. If it would form a cycle with the arcs already selected, reject it.
b. If it does not form a cycle, then add it to the tree.
If there is a choice of equal arcs, consider each in turn.
4. Repeat steps 3 until all vertices are connected.
Figure 1.1
Figure 1.1 shows Kruskal’s Algorithm in function. Once the arcs are sorted into ascending
order of weight, ‘CE’ and ‘AB’ are the options we have to start the tree. In this case ‘CE’ has been chosen.
Following this we choose arc which would not form a cycle and add it to the tree. The total cost of this
MST is 39.
Prim’s algorithm was chosen over Kruskal’s for this problem due to a continuous tree being formed. The
freedom of choosing any vertex to start the tree allows the initial starting point of the milkman to be chosen
rather than always starting with the arc of the least weight.

Assumptions:
 As the average number of people per household is 2.36 and the average consumption of milk per
week is approximately 214 ml per day7 meaning the average milk consumed per household per day is
~ 500ml (just under one pint).
 The average wage for a milkman is approximately £30 000 per year while working 80 hours each
week8 = £10.40 per hour. Also the average walking speed is 3.1 mph.
 The cost of running a milk float is 12 pence per kWh with 1kWh being used per mile.
 The average speed for a milk float varies from 5-15 mph and I have used the lower bound value.
 The time taken while stopping at each house is 45 seconds.

Model 1a:
Figure 2.0 shows a basic grid layout for a street. The milk man would drive to each
0.025 miles house till the end of the road, turn the milk float around and deliver to all the
houses on the opposite side of the road. It is assumed that there is one delivery man.
Time spent = 0.81 minutes per house. Cost = £0.14 per house.

Figure 2.0

6
Population and Household Estimates for the United Kingdom, March 2011 - ONS. 2015.
7
AHDB Dairy - Datum UK Dairy Consumption. 2015. [Accessed 03 August 2015].
8
A working life: The milkman | Money | The Guardian. 2015. Available at:
http://www.theguardian.com/money/2009/oct/03/milkman-career. [Accessed 03 August 2015].

P a g e 6 | 19
Model 1b:
6.75 meters Figure 2.1shows an alternative route for delivery. This is due to the possibility of
different delivery timings. There are thousands of possible paths which could be
0.025 miles used. I have chosen a ‘zigzag’ pattern as it seems to be the second most basic
H route available. In this model the driver would be advised to cross the road. The
milkman would park at D, after which he will visit houses A-D. Following this,
the driver can move forward to the next parking space at H. From here he can
D C
8m cover the top 6 houses on foot. Time spent = 0.92 minutes per house.
A Cost = £0.15 per house.
B
Figure 2.1

Model 2a:
The reason for investigating the cul-de-sac model is that it is a layout that occurs
frequently and will require a different strategy due to the space restrictions as
0.025 miles well as the reduced number of houses. The assumptions are the same as the
previous model. The milk float will stop at the three points shown on Figure
2.2. Since the length of the cul-de-sac is > 25 meters, a turning area will be
needed. The turning circle of a milk float is ~ 8.8 meters. Time spent = 0.79
minutes per house. Cost = £0.14 per house.
6.75 meters
Figure 2.2
Model 2b:
The same model as seen above was investigated again but with two milkmen.
This was done to see if the reduction in time taken outweighs the change in
0.025 miles salary.
Time spent = 0.42 minutes per house. Cost = £0.15 per house.
These figures suggest that for a very small increase in price per house, there
has been a large decrease in the time taken. I have concluded that using two
milkmen is the most efficient way.
6.75 meters

P a g e 7 | 19
Results for milk delivery
When the grid layout is investigated with 30 houses, we can apply a general rule based on model 1a. The
change would be to add an extra milkman who covers the opposite side of the street when the car is parked.
These algorithms are later tested by applying them to real world situations.
Algorithm for a grid like street layout:
1. If the number of houses to be delivered to is an even number, then the milkman must
park adjacent to a house as seen on Figure 3.0.
2. Each milkman must carry a crate holding 3 bottles of milk when they deliver.
3. If the number of houses is > 30 then the milk float should coast at ~ 2 mph while the
milkmen deliver thus saving time on stoppages and starts.
4. Once the 6 houses are done, the driver must drive forward to the next set of 6 houses
and repeat the cycle until all houses are done.
Figure 3.0
Time spent = 0.59 minutes per house. Cost = £0.20 per house.
Algorithm for cul-de-sac street layout:
1. Park the milk float such that there are 5-6 houses on the same side of the road which are accessible
within 5 meters.
2.
a. If a sector has an odd number of houses, the passenger delivery man
delivers to n houses and the driver must deliver to n-1 houses. This is
because the driver takes a few seconds longer to turn the car on/off
and to walk round the car to the kerb. It will allow both milkmen to
return to the milk float at approximately the same time.
b. If the sector has an even number of houses then both deliver to the
same amount. Figure 3.1
3. When at the end of the cul-de-sac, the milk float must be stopped while turning such that it is
horizontal allowing nearby houses to be reached only if there are houses at the end of the street.
4. Each milkman must carry either 2/3 bottles in a crate and replace the empty ones
5. Repeat steps 2 onwards until all deliveries are complete.
Time spent = 0.58 minutes per house. Cost = £0.17 per house.
The following models use the information obtained from the basic models along with the algorithms
researched to investigate the most efficient path for the milkmen. I am now looking at real street layouts to
demonstrate how the initial ideas
Model 3a:
The path in blue on Figure 3.2 shows the minimum spanning tree which
equals 403 meters. Deliveries have to be made in a certain order which
limits the use of algorithms due to intuition and logic being more useful.
For instance, there are two main ways of travelling from a -> b, shown in
red and yellow. The chosen route was the yellow one due to it being 9
meters shorter. However the path from b -> c is the shorter one.
Time spent = 1.18 minutes per house. Cost = £0.21 per house.

Figure 3.2
P a g e 8 | 19
Model 3b:
Figure 3.3 shows a central point, ‘M’, which is known as the centre of
minimum distance. “This method finds the location that minimizes the
combined travel distance from a set of points on a spherical earth.” Using
Prim’s algorithm, the MST would be 458 meters up to the last house. “If a
circuit visits every vertex exactly once and ends at the starting point, it
becomes known as a Hamiltonian circuit (or Hamiltonian cycle).”9 This implies
that the path in Figure 3.3 is Hamiltonian.
Time spent = 1.67 minutes per house. Cost = £0.29 per house.
Model 4: Figure 3.3

The driver would park at the green circle on Figure 4.0 and walk to
deliver to the houses. These houses were chosen using a randomiser.
In addition, as the driver is free to walk, we are able to use Prim’s
algorithm in order to find the minimum spanning tree which is 86
meters. The weights attached to each path (seen by the text in white)
equal the distance from one node to the next in meters.
Time spent = 1.15 minutes per house. Cost = £0.20 per house.
Figure 4.0

Figure 4
Model 5:
The next stage in developing the model was to
incorporate both a cul-de-sac and grid-like street
layout within one milk run thus making the model
more similar to reality. The street layout seen in
Figure 5 is of a small neighbourhood close to a local
dairy farm. Three small “sectors” were created and
Prim’s algorithm was used in order to find the most
efficient path. The main assumption is that there is
one delivery man who decides to drive to each
sector, park the milk float and deliver by foot.
Time spent = 0.97 minutes per house. Cost =
£0.13 per house.

Figure 5
9
UK Essays. November 2013. Shortest Path And Combinatorial Optimization Problems Computer Science
Essay. Available from: http://www.ukessays.com/essays/computer-science/shortest-path-and-
combinatorial-optimization-problems-computer-science-essay.php?cref=1

P a g e 9 | 19
Discussion of Results
Testing algorithm (grid layout) in a real world situation:
Figure 6 shows a highly populated street with 26 houses. The steps of the algorithm must be
followed as shown below:
1. The milk man would park on the left hand side of the street such that there is one house
adjacent on both sides to where the milk float is parked.
2. 3 bottles are carried in a crate.
3. There are < 30 houses therefore the driver can afford to park rather than let the vehicle
coast.
4. Once the two milkmen cover the 6 houses, they move on to the next set as shown by the
brackets and so forth.

Testing algorithm (cul-de-sac) in a real world situation: Figure 6


Figure 7 shows a cul-de-sac which has been split into 4 sectors. The algorithm
must be applied to test its functionality.
1. The driver would park outside the central house of each sector.
2. Since there are 5 houses, the driver would deliver to two of them and the
other milkmen would deliver to 3.
3. There are no houses at the end of the cul-de-sac however there would be
a stoppage for the second ‘sector’ before turning.
4. Repeat step 2 for the third sector.
5. The fourth sector has 4 houses therefore each delivery man will deliver
to two houses.

Limitations of these algorithms:


 I have assumed that every house in a cul-de-sac will have milk delivered
but in reality this is very unlikely because only “10% of liquid milk is
delivered to consumers via doorstep deliveries."10 This implies that
there is a very small chance of all houses in a street having milk delivered. Figure 7
 The algorithm has not been tested thoroughly and does not account for factors such as seasonal
changes and cancellation in orders.

Welcome to Bottle 2 Bottle. 2015. Available at: http://www.bottle2bottle.com/ff-milk.asp. [Accessed


10

11 August 2015].

P a g e 10 | 19
Methods and Materials (watering a field)
The second mathematical model consists of investigating the placement of sprinkler systems to water a
cricket field optimally with minimal inconvenience. It is important to cover the maximum area of the
ground but to also keep costs low and reduce the inconvenience caused to players and spectators.
Table 1.0 shows information11 researched about different sprinklers:
Name: Radius Cost (£): Sector:
(meters):
Hunter I-90 Pop up Sprinkler 30 141 Full circle and adjustable.
Rainbird 950 Pop-up with Swing Joint 28 327 Full circle
Hunter I-40 Pop up Sprinkler 13.4-21.6 76 Adjustable – 50 to 360 degrees
Hunter I-40 Pop up Sprinkler 13.4-21.6 87 Full Circle - twin opposing
water jets
Hunter G990 23.2–29.9 207 Full Circle
Hunter G995 23.2–29.9 263 Adjustable
Table 1.0
A basic model of a circular lawn which required watering will allow me to use mathematical techniques to
calculate the correct placement. This model will eventually lead to finding the best placement for
Edgbaston Cricket Ground in Birmingham and so the
measurement of the diameter replicates this ground. The
Hunter G990 was chosen as the only sprinkler for this
model. Scale:
Model 6: 1 cm = 14 meters
The mathematical application to this model is
mainly seen with relation to the spacing of the
circles and the three rows. An equilateral triangle
is formed when you take a set of 3 circles and join Minimum radius
their centres together as seen on Figure 8.1. The = 1.66 cm = 23.2
side lengths are two radii long. By using the meters
Pythagorean Theorem, I calculated the red dotted
line ~ 40.2 meters. The conclusion gain from this
is that each sprinkler should be placed 23.2 meters 23.2
23.2
apart (2 radii) and each row of sprinklers should be
placed 40.2 meters away. This approach was obtained
from an online resource that had a similar problem of
packing circles but into a rectangle rather than a circle.12
Figure 8.1
11
Pop-up sprinklers for sports pitches, golf courses, football fields and rugby pitches. 2015. Available
at: http://www.access-irrigation.co.uk/shop/sports-irrigation/sports-sprinklers. [Accessed 03 August
2015].
12
How Can We Water All Of The Grass? | Robert Kaplinsky - Glenrock Consulting, LLC.
2015. Available at: http://robertkaplinsky.com/work/water-grass/. [Accessed 03 August 2015]

P a g e 11 | 19
In addition, this arrangement of circles is known to be the most
optimum with 7 circles which don’t exceed the boundary of the
field according to a Wikipedia article13. An online calculator further
proves this, providing the area wasted as well = 22.84%. This
arrangement was then further confirmed by another source.14
Total cost = £1449 but only ~3770 pi is covered ∴ = £0.12 per
m^2.
Limitations with this basic model:
Quantity – placing 7 sprinklers in permanent positons might be too
many as players would need to train and play on the ground
meaning ‘bumps’ in the outfield could lead to injuries.
Gaps – using this method leads to many gaps in the field which need
to be covered. This problem will ruin parts of the pitch and needs
to be addressed.
Figure 8.2

13
Circle packing in a circle - Wikipedia, the free encyclopaedia. 2015. Available
at: https://en.wikipedia.org/wiki/Circle_packing_in_a_circle. [Accessed 04 August 2015].
The best known packings of equal circles in a circle. 2015. Available at:http://hydra.nat.uni-
14

magdeburg.de/packing/cci/cci.html#Applications. [Accessed 04 August 2015].

P a g e 12 | 19
Results (watering a field)
Model 7:
The advancement from the first model would be to add a rectangular box which would act as the wicket.
This would need to be avoided in the radius of the sprinklers and would have its own method of watering.
The application of the algorithm below is seen in Figure 9.0
Algorithmic approach:
1. Cover as much of the ground as possible with sprinklers making sure no radii overlap:
a. Place one sprinkler in the south west corner such that its edge is touching the boundary of the
field. Not to scale
b. Place another sprinkler adjacent to it to the right without the
radii overlapping but with its edge in contact with the boundary.
c. Place another sprinkler directly above the point where the radii
of the two sprinklers meet.
d. Form an equilateral triangle, whose side length should be two
radii long, using the point where each sprinkler is placed.
e. The vertical distance from the intersection point to the 3rd
sprinkler = the distance between the rows.
f. The side length = how far apart each sprinkler should be placed.
2. Use half the region (semi circles) of the sprinklers and place them so
there are no overlaps.
3. If there are still parts of the field which aren’t watered, then shift
Figure 9.0
the sprinklers to cover the gap making sure there is the least possible overlap.
4. If there are any regions which don’t need to be watered, the sprinkler under whose domain this is in
should be discarded.
Applying the algorithm to Figure 9.0:
Total area to cover = (4900 pi – 603.6) m^2.
Total cost = (6*207) + (8*76) = £1850 which covers ~12861 meters ∴ =
£0.140 per m^2.

Model 8a:
Some cricket fields around the world have a rectangular shape with a
curvature. For instance Lords in England and the Brabourne stadium in India.
This led to the development of a strategy to find the optimum placement of
sprinklers for such a field. (Figure 9.1)
The total area is given by the formula, Area = LW - 4r2 + πr2. Where W =
overall width, L = length and r= curvature of the corner. This gives a total
area of 20156.03 m2.
Figure 9.1
When the algorithm from model 6 and the Pythagorean Theorem is applied, the total area
covered was 16991.68 m2. The total cost in this situation would be £2052 = £0.121 per m2.

P a g e 13 | 19
Model 8b:
Another strategy developed to cover the maximum area possible can be seen
opposite. The criterion for this model was to only use the maximum radii
sprinklers. The total area covered by this arrangement was 5400pi =
16964.6 m2. The total cost = £1656 = £0.098 per m2. In comparison to
the previous model, the area covered is slightly less but both the total cost
and the cost per m2 is considerably less. (Figure 9.2)

Algorithmic approach:
1. Set two Hunter G990 sprinklers at their maximum radii such that
they are in contact with the edges adjacent to the top or bottom
corners. Figure 9.2
2. Place two sprinklers of appropriate radii above/below the two which are already in place so that
their outer edges are in contact.
a. If there is an area which has to be avoided in the radii of these two sprinklers then reduce
the radii.
3. Choose a sprinkler with an appropriate radius in order to fill the space in between these four
sprinklers.
4. Set the radii of one/many sprinkler/s so that each will fully cover the remaining corners.
5. Choose a sprinkler with an appropriate radius in order to fill the space in between this sprinkler/s
and the central ones. Note: this step can be ignored if the placement
of a sprinkler would lead to an overlap.
6. Any remaining gaps which can be covered using the minimum
radius sprinkler should be covered as long as there is minimal
overlap.
Applying the algorithm to Figure 9.3:
The total area to cover = 20156.03 – 1340.64 = 18815.39.
The total area covered = 16659.08 m2 therefore the area not covered ~
11.5%.
Cost: £1588 ∴ = £0.095 per m2.

Figure 9.3

P a g e 14 | 19
Discussion of Results (watering a field)
The algorithmic approach to model 7 has numerous limitations:
 Spectators – the radii of some sprinklers are still exceeding the edge of the field which could pose a
possible inconvenience to the public in the stands.
 Quantity – the number of sprinklers has increased from the previous model due to maximum
coverage being the priority.
 The 13% of space uncovered can be major problem for a potential client.
Due to this, overlaps must be considered.
The change to the algorithm developed would be to step 2. The new
instruction would read as follows:
“2. Use half the region (semi circles) of the sprinklers and place them so all
gaps are covered while minimizing the size of the overlap.”
This change can be seen on Figure 10.0 as the radii of the pre-existing black
semi-circles have been increased to water the 13% of the ground which
was previously uncovered. The change in cost would be the additional two
sprinklers adjacent to the central wicket. These can be seen by the blue
outline. The new total cost would be £2002 ∴ = £0.135 per m^2.

Similar to the previous model, the client might not be happy with 11.5% of the ground
Figure 10.0
not being watered at Lord’s (model 8b). The changes can be seen in Figure.
The changes to the algorithm would be as follows:
“5. Choose a sprinkler with an appropriate radius and region shape (full/half)
in order to fill the space in between the sprinkler/s and the central ones. Allow
an overlap if all the space is not covered.
6. Any remaining gaps which can be covered using an appropriate radius
sprinkler should be covered. Allow an overlap if all the space is not covered.”
Calculations:
The eight purple sprinklers have been upgraded from the ones that cost £76 and
have a maximum radius of 21.6 meters to those that cost £207 with a maximum
radius of 29.9 m. The new total cost = £2636 ∴ = £0.140 per m2.

There would still be limitations with these changes: Figure 10.1


 The main drawback with this model is the overlaps caused which might result in over watering of
parts of the ground.
 Furthermore, the edges of the ground are slightly exceeded which could cause problems with the
crowd and staff when watering the ground during an innings change.

P a g e 15 | 19
Conclusions – Milk Delivery
Having investigated and analysed numerous models, I have come to the following conclusions:
 When the delivery run is to a grid street layout, the most efficient method is of using two delivery
men who complete the delivery simultaneously. This is best done in a sector of 6 houses as each milk
man can deliver to three each. This model produced the least time for delivery at 0.59 minutes per
house. However this method is not the cheapest at £0.20 per house. There is a payoff between the
cheapest method and the quickest method meaning that the ‘best’ model would depend on the
individual’s interpretation of the phrase an ‘efficient delivery.’
 The cheapest individual delivery method to a grid layout (not including the neighbourhood) was the
initial model 1a. It was a simple model consisting of one milkman who would stop at each house.
This meant the time taken per house was one of the longest at 0.81 minutes.
 Having explored two strategies for delivering milk to a cul-de-sac, the most efficient method is
outlined in the algorithm developed. Delivery of two milkmen on foot produced figures of 1.15
minutes and £0.20 per house. The figures for model 2a (0.79 minutes and £0.14 per house) suggest
that a combination of both walking and driving result in the shortest and least expensive technique.

Conclusions – Watering a cricket field


Having investigated and analysed numerous models, I have come to the following conclusions:
 The use of trigonometry in order to find the optimum placement of sprinklers produced two options.
Firstly, it is confirmed that the placement of seven sprinklers in model 6 does cover the maximum
area possible without overlaps optimally. Similar to the previous problem there would be a payoff. In
this case it would be between the area covered and the amount of overlap. Secondly, the area not
covered by the full radii of sprinklers can be fulfilled by using half the region of numerous smaller
sprinklers. This will result in the total cost increasing but would reduce the cost per m 2 as more of
the ground is covered.

P a g e 16 | 19
Critical Evaluation and scope for further research
If the work to find the most efficient method for milk delivery were to continue, it would be best
explored through the awareness of other factors such as weather and traffic. As well as through practical
experiments in further testing of the mathematical models produced. There are numerous real world effects
which can’t be accounted for and analysed through models. The figures used for lengths of roads do have a
degree of inaccuracy as they are obtained from Google Maps. The viability of these figures could be
improved through real world measurements where the theoretical approach can be approached practically.
With regards to the watering of the cricket field, further research on the specifics of a sprinkler
system such as storage units, volume of water required and costs of the individual equipment would provide
a more thorough analysis of the problem. The models in this report present a summarised study of the
situation and this can be further explored through practical experiments. For instance visiting nearby cricket
grounds and speaking to groundsman would help us to understand the unexpected factors which aren’t
explored in mathematical models such as weather conditions and timing restrictions.
Furthermore, I only considered the cricket fields in the UK and did not look at other countries where
there might be slightly different shapes and watering requirements. For example the weather in India and
the uncertainty of the water supply in parts of the country means the time and costs of sprinkler systems
would differ. This in turn might affect their placement and require a different strategy.

P a g e 17 | 19
Bibliography
A working life: The milkman | Money | The Guardian. 2015. Available at:
http://www.theguardian.com/money/2009/oct/03/milkman-career.
AHDB Dairy - Datum UK Dairy Consumption. 2015. Available at: http://dairy.ahdb.org.uk/market-
information/dairy-sales-consumption/uk-dairy-consumption/#.VbnrR_lVhBc.
BBC Sport Academy | Cricket | Rules | The basics | Pitch dimensions. 2015 Available at:
http://news.bbc.co.uk/sportacademy/hi/sa/cricket/rules/the_basics/newsid_3795000/3795261.stm.
[Accessed 03 August 2015].
Circle packing in a circle - Wikipedia, the free encyclopedia. 2015. Available
at: https://en.wikipedia.org/wiki/Circle_packing_in_a_circle. [Accessed 04 August 2015].
Highways, transportation and development - Part 3 - Highway, Transportation And Development. 2015
Available at: http://www.leics.gov.uk/index/htd/highway_req_development_part3.htm.
How Can We Water All Of The Grass? | Robert Kaplinsky - Glenrock Consulting, LLC. 2015. Available
at:http://robertkaplinsky.com/work/water-grass/. [Accessed 03 August 2015]
Jameson, S. (2010). Edexcel AS and A Level Modular Mathematics Decision Mathematics 1 D1. Edition:
Edexcel Limited
LS Irrigation - Cricket Pitch Irrigation Systems & Services. 2015. Available
at: http://www.lsirrigation.co.uk/sports-irrigation/cricket-pitch-watering-irrigation/cricket-pitch-
irrigation.html. [Accessed 03 August 2015].
Pop-up sprinklers for sports pitches, golf courses, football fields and rugby pitches. 2015. Available
at: http://www.access-irrigation.co.uk/shop/sports-irrigation/sports-sprinklers. [Accessed 03 August
2015].
Population and Household Estimates for the United Kingdom, March 2011 - ONS . 2015. Available
at:http://www.ons.gov.uk/ons/rel/census/2011-census/population-estimates-by-five-year-age-bands--
and-household-estimates--for-local-authorities-in-the-united-kingdom/stb-population-and-household-
estimates-for-the-united-kingdom-march-2011.html.
The best known packings of equal circles in a circle. 2015. Available at:http://hydra.nat.uni-
magdeburg.de/packing/cci/cci.html#Applications. [Accessed 04 August 2015].
UK Essays. November 2013. Shortest Path And Combinatorial Optimization Problems Computer Science
Essay. Available from: http://www.ukessays.com/essays/computer-science/shortest-path-and-
combinatorial-optimization-problems-computer-science-essay.php?cref=1
UCL - D1_graphs_glossary Available at:
http://www.ucl.ac.uk/~ucahbtw/docs/d1lesson1/D1_graphs_glossary.pdf [Accessed 11 August 2015].

P a g e 18 | 19
Acknowledgements
I would like to thank Dr. Ewan Russell for his excellent guidance, patience, and providing me with an excellent topic
for research. I am extremely grateful for the ongoing support and advice provided throughout the
placement.

I would also like to thank Dr Farzana Aslam and Mr Steve Joiner for making the placement a possibility and
providing this opportunity.

P a g e 19 | 19
Appendix 3:
Calculations for model 1a:
 Time to travel 0.05 miles (2x0.025 miles) = (0.05/5)*60*60 = 0.6 minutes.
 Distance driven = 0.05 miles = 0.6 pence
 Time spent stopping at houses = 10*0.75 minutes = 7.5 minutes.
 Total time spent = 8.1 minutes for 10 houses= 0.81 minutes per house.
 Average wage for milkmen = £30 000 per year @ 80 hours a week so ~ £10.40/hour.
 Milkmen usually start work at 2:45 AM and finish around 12 pm (Monday to Saturday) = 9 ¼ hours.
This means that the ~ cost for the 8.1 minutes it takes to deliver to 10 houses is £1.40 therefore the
(~) total cost to the client would be £1.41 = £0.14 per house.
Calculations for model 1b:
 3 miles/hr walking speed = 83 meters/minute.
 The minimum carriageway width is 6.75m15.
 Total walking distance = 62.375 meters ∴ total time ~ 46.2 seconds
 Distance driven = 0.05 miles = 0.6 pence = 59 seconds.
 Time Spent at each house = 45 seconds ∴ = 7.5 minutes.
 Total time spent = 9.25 minutes ∴ ~ 0.92 minutes per house
 Total cost (wages + electricity) = 1.45+0.006 = £1.46 ∴ ~ £0.15 per house.
Calculations for model 2a:
 Total distance driven = 2 * 0.025 miles + 8.8 meters = 89.3 meters
 Time spent at houses = 15*0.75 = 11.25 minutes.
 Time spent driving ~ 0.67 minutes∴ ~ total time = 11.9 minutes = 0.79 minutes per house.
 Wages cost ~ £2.06 and petrol costs = 0.67 pence ∴ ~ total cost = £2.07 = £0.14 per house.
Calculations for model 2b:
 Total distance driven = 2 * 0.025 miles + 8.8 meters = 89.3 meters
 Time spent at houses = (15*0.75)/2 = 5.625 minutes.
 Time spent driving ~ 0.67 minutes∴ ~ total time = 6.275 minutes = 0.42 minutes per house.
 Wages cost ~ £1.09 *2 = £2.18 and petrol costs = 0.67 pence ∴ ~ total cost = £2.18= £0.15
per house.

Highways, transportation and development - Part 3 - Highway, Transportation And Development.


15

2015. Available at: http://www.leics.gov.uk/index/htd/highway_req_development_part3.htm.


Calculations for grid layout algorithm:
 Total distance driven = 12 meters * 5 ‘sectors’ = 60 meters ∴ cost = 0.45 pence
 Total distance walked = 36.75 * 5 = 183.75 meters
 Time spent at each house = 30 seconds as each milkman sorts out 3 orders at when he delivers ∴ =
15 minutes
 Time to walk = 2.21 minutes.
 Time to drive = 0.45 minutes ∴ ~ total time = 17.7 minutes = 0.59 minutes per house
 Wages cost = £6.14 (two milkmen) ∴ total cost = £6.14 = £0.20 per house
Calculations for cul-de-sac layout algorithm:
 Distance walked for one ‘sector’ = 8+11+11+19+19 (distances obtained using google maps) = 68
meters. = 204 meters for 3 sectors.
 Total distance driven = 2 * length + turning circle = 2 * 0.025 miles + 8.8 meters = 89.26 meters.
 Time at houses = (0.75 * 15)/2 = 5.625 minutes. Time spent driving = 0.65 minutes. Time spent
walking = 2.46 minutes ∴ = 8.74 minutes overall ~ 0.58 minutes per house.
 Fuel cost = £0.00648. Wages = £2.49 = £0.17 per house.
Calculations for model 2a:
 Total distance driven = 2 * 0.025 miles + 8.8 meters = 89.3 meters
 Time spent at houses = 15*0.75 = 11.25 minutes.
 Time spent driving ~ 0.67 minutes∴ ~ total time = 11.9 minutes = 0.79 minutes per house.
 Wages cost ~ £2.06 and petrol costs = 0.67 pence ∴ ~ total cost = £2.07 = £0.14 per house.
Calculations for model 3a:
 Time to travel by car = 403 meters. 5mph = 8047 meters per hour ∴ ~ 3 minutes.
 Time spent stopping at houses = 7*0.75 minutes = 5.25 minutes.
 Total time = 8.25 minutes for 7 houses ∴ ~ 1.18 minutes per house
 Distance driven = 403 meters = 3 pence
 Wages cost = £1.43 ∴ total cost = £1.46 for 7 houses ∴ ~ £0.21 per house
Calculations for model 3b:
 Distance travelled by car = 0 meters as it is parked therefore saving 3 pence from model a.
 Time spent stopping at houses = 7*0.75 minutes = 5.25 minutes.
 Time spent walking = 533 meters @ 83 meters a minute = 6.42 minutes.
 Total time spent = 11.7 minutes ∴ ~ wages = £2.03 for 7 houses ∴ ~ £0.29 per house.
 Time = 1.67 minutes per house.
Calculations for model 4:
 Time spent stopping at houses = 4*0.75 minutes = 3 minutes.
 Total walking distance = 134 meters = 1.61 minutes
 Total time spent = 4.61 minutes for 4 houses ∴ ~ 1.15 minutes per house
 Total cost = £0.80 for 4 houses ∴ ~ £0.20 per house
Calculations for model 5:
 Time spent stopping at houses = 14*0.75 minutes = 10.25 minutes.
 Total distance walked = 282 meters = 3.4 minutes
 Total time = 13.6 minutes for 14 houses = 0.97 minutes per house
 Total distance driven = 256 meters = 1.91 pence
 Wages cost = £1.78 ∴ total cost =~ £1.80 for 14 houses = £0.13 per house.

You might also like