101 ISSN 2086-5953
ANT COLONY ALGORITHM BASED GENERIC MODEL FOR SUPPLY CHAIN PERFORMANCE MEASUREMENT
Taufik Djatna
1
, Ditdit N. Utama
2 1
Postgraduate Program of Agro-Industrial Technology, Graduate School of Bogor Agricultural University, Darmaga P.O. Box 220 Bogor 16002, Indonesia
2
Information System Department, Science and Technology Faculty of State Islamic University Syarif Hidayatullah
– Jakarta, Ir. H. Juanda Street No. 95 Ciputat, Jakarta, Indonesia E-Mail: taufikdjatnaipb.ac.id
1
, ditditnhotmail.com
2
ABSTRACT
Ant colony algorithm is one of many mature research methodologies in searching, routing and
optimizing field. This methodology uses biological and behavior phenomenon as an algorithm
philosophy to solve fastest path problem and other optimization problems. Many researchers have
researched by using ant colony algorithm as their methodology, not only for technical cases, but also
for management aspect cases. The ant colony algorithm methodology can be used in problem
solving for the routing or optimizing supply chain path in supply chain management case. As we
know, that in the supply chain case, the best path searching problem is a usual problem happen. In
this paper, the authors propose a generic model based on ant colony algorithm that was
implemented
in supply
chain performance
measurement cases. Keywords: ant colony algorithm, generic model,
supply chain, performance measurement
1
INTRODUCTION 1.1
Background
More than one company involvement in industrial sector is the reason supply chain
management concept is needed in industrial implementation. Relationship among companies in
one industrial area, such as: raw material producers, collectors, factories, warehouses, distributors,
retailers, and the customers, is a supply chain flow that could not be separated one to each others. The
differentiation of data and information platform, leadership style, and the working culture in every
level of supply chain, is the problem that has to be solved.
Supply chain is a system where the organizations or companies can supply their
products and services to their customers. This chain is as network of many related companies that have
the same goals and objectives. The supply term, in supply chain term, includes product transformation,
in this case, transforming raw material to end product [1].
On the other hand, the best path of supply chain searching problem, to supply the product
from supply to demand side [2], always be researched in day per day by many researchers. In
other research, [3] has researched in supply chain management as Intelligent Supply Chain. ISN is a
system that integrates all of the involved actors in supply chain. [4] has researched in responsive and
efficient supply chain.
One of many methodologies that can be used for solving the current supply chain problem is ant
colony algorithm methodology, or usually called Ant Colony Optimization ACO. ACO is a
paradigm for designing meta-heuristic algorithms for combinatorial optimization problems. The
essential trait of ACO algorithms is the combination of a priori information about the
structure of a promising solution with a posteriori information about the structure of previously
obtained good solutions [5]. Other definition said that ACO is a heuristic optimization method for the
fastest path problem solving or other problem optimization by using biological ant behavior
phenomenon [6].
The meaning of generic model in this case is the model that describes the developed model
generally. On the other hand, the generic model is the opened model. It can be extended in accordance
with what we want, which means our model will support solution in different process in the supply
chain. According to that background, this paper would explained research background, related
works, theoretical review in the first part; generic model
development in
the second
part; measurement and result in the third part; and
discussion and conclusion in the fourth part.
1.2 Related Works
There are many researches that used ant
ISSN 2086-5953 colony optimization as a main technique. Some of
them are [7] researched in production optimization and maintenance process scheduling area; [8]
researched by using ACO in routing design area; [9] have researched in project critical path
searching area by using ACO as a main method [10] also studied in product packing and delivering
area for supply chain; and [11] have studied aspect of supply chain in performance management, in
combination of SCOR, value added, cost and financial supply chain performance measurement.
On the other hand, [12] combined ACO and Genetic Algorithm for developing the Chinese
question and answering system. [13] used ACO for data processing dynamic city choice case and [14]
used ACO for partner selection case. Moreover, [15] and [16] used ACO that was implemented in
mobile system case.
1.3 Theoretical Review
Supply Chain Management
Supply chain concept is still a growing concept in logistic problem view. In a conventional
concept, logistic problem has been seen as internal company problem only. The companies have to
solve their problems internally and solely, without collaborate to their company partners. In this new
one, logistic problem is seen as a boarder problem. The logistic problem can be happen in every step of
supply chain, from the first level up to the last level of supply chain. That is reason why supply chain
management can be defined as a set of approaches utilized
to efficiently
integrate suppliers,
manufactures, warehouse, and stores, so that merchandise is produced and distributed at the right
quantities, to the right locations, at the right time, in order to minimize system wide costs while
satisfying service level requirement [2]. The figure 1 described supply chain management concept [2].
Figure 1. Concept of Supply Chain Management
On the other hand, supply chain can be defined as a system that used by company to supply
its product and service to its customers. This chain is also as network of many related organizations
that have the same goals and objectives [1].
Ant Colony Optimization
[17] Said that ACO is an optimization technique inspired by observations on the behavior
of biological ant colonies; it has been introduced by [13], and [19]; and developed later into a meta-
heuristic for combinatorial optimization problems [20], [21], [22]. Recently, ACO algorithms have
also turned out as competitive in a discrete stochastic optimization context [23], [24], [25],
[26], [27].
An ACO is essentially a system based on agent, which simulates the natural behavior of ants
including mechanisms
of co-operation
and adaptation. To use this kind of system, a new meta-
heuristic is proposed in order to solve combinatorial optimization problems. This new meta-heuristic has
been shown to be both robust and versatile in the sense that it has been successfully applied to
arrange of different combinatorial optimization problems. ACO algorithms are based on following
ideas [28]: 1.
Each path, is followed by an ant, is associated with a candidate solution for a given problem.
2. When an ant follows a path, the amount of
pheromone deposited on that path. This path is proportional to the quality of the corresponding
candidate solution for the target problem. 3.
When an ant has to choose between two and more paths, the path with a larger amount of
pheromone has a greater probability of being chosen by the ant.
Conceptually, in ant colony algorithm, two ways of food searching way, that are used by ants,
are used in this method. Other principles of ACO are [29]:
1. Random direction for searching and collecting
some food, 2.
When one ant finds food resource, that ant traces back again the path that was used by it,
3. The ants put chemical Pheromones on their
path that can be evaporated, 4.
In loop, the chemical pheromones will decreases, but on the other hand, it will be
added, 5.
Some ants will follow the shortest path, the strongest chemical pheromone path.
2 GENERIC MODEL
Based on basic behavior of ACO model [22], we generate a generic mathematical model for
supply chain performance measurement which was described in the next formula.
ISSN 2086-5953 Where:
= Performance of Current Path =
Current Distance,
the basis
of measurement
= Performance of Previous Node Variable = Total of Previous Node Variable
= Current Path Pheromone The supply chain performance depends on
the path length in this case, distance is the basis variable for measurement, supply chain element
performance and pheromone quality. The supply chain element performance could be showed in the
next mathematic formula.
Where: = Performance of previous element
= Performance of prev ious element‘s
SCOR variable = Performance of Previous
element‘s Financial variable
= Performance of Previous element‘s
Added Value Variable =
Performance of previous element‘s Transportation Cost Variable
The supply chain element performance depends on performance of SCOR, financial,
added-value and transportation cost variables. This is an opened model, so if it is needed, we can add
some new variables, according to the case that was being faced.
The generic model can be described in usecase diagram. This model is divided in to two
sub-systems; Ant colony sub-system and Supply Chain Performance Sub-System. In the ant colony
sub-system, user can use the main model to search the best path of supply chain. On the other hand,
supply chain and supply chain element performance are measured in supply chain performance sub-
system. The usecase diagram could be showed in Figure 2.
Information flow and type of the model could be described by using information liquid
model Figure 3. This model could be showed some information resources that were needed in this
model, such as: supply chain element; value of SCOR Based on [30], financial, added value and
transportation cost variable; and distance and route.
The generic model is shown in class diagram Figure 4. To describe the model, there are five
classes in this class diagram: class AntGraph, class Ant, class AntColony, class Ant4SC and class
AntColony4SC. Class AntGraph connects to ClassSupplyChainPerformance that consists of:
class SCORPerformance,
class financialPerformance,
costPerformance, and
addedValuePerformance. Each class represents the name of main component in ACO. This model can
be called by distance based model [11].
Ant Colony Sub-System
Supply Chain Performance Sub-System User
The Best SC Searching
Criteria Identification
SCOR Performance
Added Value Variable
Performance Financial
Variable Performance
SC Element Performance
SC Performance
Transportation Cost Variable
Performance
Figure 2. Usecase Diagram of Generic Model
Figure 3. Information Liquid Model
ISSN 2086-5953
0..1 0..
0..1 0..
antGraph 0..1
0..
0..1 0..
0..1 0..
0..1 0..
0..1 0..
0..1 0..
0..1 0..
0..1 0..
ant -
- -
- -
- -
- -
- -
m_nAntID m_path
m_nCurNode m_nStartNode
m_dPathValue m_pathVect
s_nAntIDCounter s_dBestPathValue
s_bestPathVect s_bestPath
s_nLastBestPathIteration : int
: int[][] : int
: int : double
: int : int
: double : int
: int[][] : int
+ +
+ +
+ +
+ +
+ +
setAntColony reset
ant int nStartNode, Observer observer init
start run
better double dPathValue, double dBestPathValue
stateTransitionRule int r localUpdatingRule int r, int s
end getBestPath
: void : void
: void : void
: void : boolean
: int : void
: boolean : int[]
antColony -
- -
- -
m_ants m_nAnts
m_nAntCounter m_nIterCounter
Attribute_5 : int
: int : int
: int : int
+ +
+ +
+ +
+ +
+ +
+ +
+ antColony int nAnts, int nIterations
createAnts int ants getBestPathVector
start iteration
getGraph getAnts
getIterations getIterationCounter
getID update Obeservable ant, Object obj
getBestPathValue getBestPath
getLastBestPathIteration done
globalUpdatingRule : Ant[]
: vector : int
: void : int
: int : int
: int : void
: double : int[]
: int : boolean
: void AntGraph
- -
- -
m_delta m_tau
m_nNodes m_nTau
: double[][] : double[][]
: int : double
+ +
+ +
+ +
+ +
+ +
antGraph int nNodes, double[] delta[], double[] tau[]
delta int r, int s tau int r, int s
etha int r, int s nodes
tauD updateTau int r, int s, double value
averageDelta averageTau
average double[] matrix[] : double
: double : double
: int : double
: void : double
: double : double
supplyChainPerformance -
- IDPerformance
SCPerformance : String
: double +
+ dataLoad
supplyChainPerformanceMeasurement : double
: double
Distance -
- IDDistance
distance : String
: double +
+ distanceMeasurement
distanceLoad : double
: double SCORPerformance
- -
IDSCOR SCORValue
: String : double
+ +
SCORValueMeasurement dataView
: double : void
financialPerformance -
- IDFinancial
financialPerformanceValue : String
: double +
+ financialPerformanceMeasurement
dataView : double
: void costPerformance
- -
IDCost CostPerformanceValue
: String : double
+ +
costPerformanceMeasurement dataView
: double : void
addedValuePerformance -
- IDAddedValue
addedValuePerformance : string
: double +
+ addedValuePerformanceMeasurement
dataVeiw : void
: void
DistPar -
- Distance
ParentVert : double
: int + DistPar : void
Vertex -
- Label
isInTree : char
: boolean + Vertex : void
Figure 4. Class Diagram of Model
The detailed ant colony based generic model could be described in Figure 5. It is started at
variable initiations, ants generating, up to supply chain performance value getting.
Further explanation of this model could be described in sequence diagram in Figure 6. This
diagram explained the relationship among objects or classes in the model. In the other words, this
diagram explained the relationship among entities in the model.
SequenceDiagram_1
PoCost PoAddedValue
PoFinancial PoSCOR
SumOfAnt pheromones, Evaporation
OptimumPath call
call call
call graph
call call
graph graph
generate Actor_1
Graph Ant
AntColony DistPar
Vertex SupplyChainPerformance
SCORPerformance FinancialPerformance
CostPerformance AddedValuePerformance
PoCost PoAddedValue
PoFinancial PoSCOR
SumOfAnt pheromones, Evaporation
OptimumPath call
call call
call graph
call call
graph graph
generate
Figure 6. Sequence Diagram of Model Figure 5. Activity Diagram of Model
ISSN 2086-5953
3 MEASUREMENT AND RESULT
In order to justify our model, we do some measurement based on Figure 4. Data and
information resource for this step is derived from field observation and computation.
We verify the ACO generic model on an example for this measurement which is used for
supply chain performance measurement case was described in a tree view in Figure 7. We deploy a
spanning tree algorithm to construct this tree.
The authors take this example from palm based bio-energy supply chain case. The supply
chain network described that A is the end customer; B and C are bio-energy factories; D is a Crude
Palm Oil CPO factory; and E, F and G are palm farmers.
The distance value of the tree was converted in to graph matrix in Table 1. This matrix described
distance between nodes or elements in one supply chain network.
A B
C D
F E
G 55
75 40
55 50
45 35
Figure 7. Supply Chain Tree Table 1. Graph Matrix
A B
C D
E F
G A
0.00 55.0
75.00 0.00
0.00 0.00
0.00
B 0.00
0.00 0.00
40.0 0.00
0.00 0.00
C 0.00
0.00 0.00
55.00 0.00
0.00 0.00
D
0.00 0.00
0.00 0.00
50.0 45.00
35.0
E 0.00
0.00 0.00
0.00 0.00
0.00 0.00
F 0.00
0.00 0.00
0.00 0.00
0.00 0.00
G 0.00
0.00 0.00
0.00 0.00
0.00 0.00
The performance value of each element: SCOR, financial, cost and added value, could be
shown by data in Table 2.
Table 2. Performance Value
A B
C D
E F
G SCOR
0.7 0.7
0.5 0,5
0.5 0.5
0.7 Financial
0.5 0.7
0.7 0.7
1.0 0.5
0.7 Cost
0.7 0.7
0.5 0.7
1.0 0.5
0.7 Added Value
0.7 0.7
0.5 1.0
0.5 0.5
0.5
The performance evaluation could be explained by using result of the model execution.
Some data resulted from this execution are in Table 3.
Table 3. Execution Result ∑Ph.
∑Ev. Path
Distance Performance Value
0.11 0.01
A,C,D,F 175
50.77 0.13
0.03 A,C,D,E
180 29.92
0.01 0.15
A,B,D,E 145
81.45 0.05
0.05 A,B,D,E
145 61.13
0.05 0.07
A,B,D,E 145
73.76
∑Ph. = Average of Pheromones, ∑Ev. = Average of Evaporation,
Sum of Ant = 2000 The execution result showed that optimum
path is A, B, D, E and the best performance value is 61.13, the most little one. On the other hand, the
best performance value is 29.92 with path: A, C, D, E.This performance value was described in km
kilometers unit size.
4 CONCLUSION AND DISCUSSION
Ant Colony
Algorithm could
be implemented in model of supply chain performance
measurement. It was called ant colony algorithm based generic model for supply chain performance
measurement.
Usecase diagram showed the relationship between user and the model, and between both sub-
systems; ant colony and supply chain performance sub-system.
This generic model is an opened model, because there are many variables that can be added
in this model, it depends on the case was being faced. In this paper case, author used 4 variables:
SCOR, financial, added value and transportation cost variable.
REFERENCES
[1] Indrajit RE, and Djokopranoto R. 2002.
Supply Chain
Management Concept.
Grasindo [2]
Levi DS, Kaminsky P, and Levi ES. 2000. Designing and Managing the Supply Chain:
Concepts, Strategies and Case Studies . Irwin
McGraw-Hill [3]
Rajamani D. 2005. Intelligent Supply Network
. The University of Texas School of Management
[4] Hugos M. 2006. Essentials of Supply Chain
Management . John Wiley Sons, Inc.
Second Edition [5]
Maniezzo V, Gambardella LM, and Luigi FD. 2004. Ant Colony Optimization. Internet
Publication [6]
Habilation LZ, and Hammer B. 2003. Ant Colony Optimization
. Osnabruck University
ISSN 2086-5953 [7]
Berrichi A, Yalaoui F, Amodeo L, and Mezghiche M. 2010. Bi-Objective Ant
Colony Optimization Approach to Optimize Production and Maintenance Scheduling.
Computers and Operations Research 37 2010: 1584
– 1596 [8]
Barcos L, Rodriguez V, Alvarez MJ, and Robuste F. 2010. Routing Design for Less-
Than-Truckload Motor Carriers Using Ant Colony
Optimization. Transportation
Research Part E 46 2010: 367 – 383
[9] Duan Q, and Liao TW. 2010. Improved Ant
Colony Optimization
Algorithms for
Determining Project Critical Paths. Journal of Automation in Construction 2010: 1
– 18
[10] Catay B. 2010. A New Saving-Based Ant Algorithm for the Vehicle Routing Problem
with Simultaneous Pickup and Delivery. Journal of Expert System with Applications
2010: 1 – 9
[11] Djatna T, and Utama DN. 2010. Modeling UML for Optimization of Supply Chain in
Palm Oil Based Bioenergy. International Conference
– AFITA 2010 [12] Di S, He P, and Li H. 2010. Research and
Implement on Genetic Algorithm and Ant Colony Algorithm in Chinese Question and
Answering System. Proc. Second Int Computer Engineering and Applications
ICCEA Conf, 2010, 1, 166-169
[13] Yun BJ, and Yong LS. 2009. Improved Ant Colony Algorithm with Emphasis on Data
Processing and Dynamic City Choice. Proc. Int. Conf. Information Engineering and
Computer Science ICIECS 2009, 2009, 1-4
[14] Wei DH. 2009. Improved Ant Colony Algorithm for Partner Selection. Proc. Int.
Conf. Management Science and Engineering ICMSE 2009, 2009, 265-270
[15] Wang JZ, and Hong W. 2009. Application of an Ant Colony Algorithm Based on Complex
Networks in Migration of Mobile Agents. Proc. IEEE Int. Symp. IT in Medicine
Education ITIME 09, 2009, 1, 821-825
[16] Jin WC, Jie WW, and Hong J. 2009. Cross Layer
Ant Colony
Algorithm for
Cooperative Relay Multicast Tree in Cellular Mobile System. Proc. Fifth Int. Conf.
Natural Computation ICNC 09, 2009, 3, 30- 34
[17] Gutjahr WJ. 2006. On the Finite-Time Dynamics of Ant Colony Optimization.
Methodol Comput Appl Probab 2006 8: 105
–133 [18] Dorigo M, Maniezzo V, and Colorni A.
1991. The Ant System: an autocatalytic optimization process. Technical Report 91-
016, Dept. of Electronics, Politecnico di Milano: Italy
[19] Dorigo M, Maniezzo V, and Colorni A. 1996. Ant System: optimization by a colony
of cooperating agents. IEEETransactionson Systems,Man, and Cybernetics, Part B vol.
261: 29 –41
[20] Dorigo M and Di Caro G. 1999. The Ant Colony Optimization metaheuristic, In D.
Corne M. Dorigo, and Glover F, eds., New Ideas in Optimization, 11
–32, McGraw Hill: London, UK
[21] Dorigo M, Di Caro G, and Gambardella LM. 1999.
Ant algorithms
for discrete
optimization, Artificial Life vol. 52: 137 –
172 [22] Dorigo M and Stutzle T. 2004. Ant Colony
Optimization, MIT Press: Cambridge, MA [23] Bianchi L, Gambardella LM, and Dorigo M.
2002. Solving the homogeneous probabilistic travelling salesman problem by the ACO
metaheuristic. Proceedingd ANTS ‘02, 3rd International Workshop on Ant Algorithms:
177 –187
[24] Gutjahr WJ. 2003. A converging ACO algorithm
for stochastic
combinatorial optimization. Proceedings SAGA 2003
Stochastic Algorithms: Foundations and Applications, In A. Albrecht and K.
Steinhoefl eds., Lecture Notes in Computer Science 2827, 10
–25, Springer: Berlin, Germany
[25] Gutjahr WJ. 2004. S-ACO: an ant-based approach to combinatorial optimization
under uncertainty, Ant Colony Optimization and Swarm Intelligence: Proceedings of
ANTS 2004
—Fourth International
Workshop, Lecture Notes in Computer Science 3172: 238
–249, Springer: Berlin, Germany
[26] Rauner M, Brailsford SC, Gutjahr WJ, and Zeppelzauer W. 2005. Optimal screening
policies for diabetic retinopathy using a combined discrete-event simulation and ant
colony optimization approach. In J. G. Anderson and M. Katzper eds., Proceeding
International Conference on Health Sciences
Simulation, Western MultiConference ‘05: 147
–152 [27] Birattari M, Balaprakash P, and Dorigo M.
2005. ACO
F-Race: Ant
colony optimization and racing techniques for
combinatorial optimization
under
ISSN 2086-5953 uncertainty. Proceedings of MIC 2005
—6
th
Metaheuristics International
Conference: 107
–112 [28] Sinha AN, Das N, and Sahoo G. 2007. Ant
Colony Based Hybrid Optimization for Data Clustering. Kybernetes Vol. 36 No. 2, 2007:
175-191 [29] Ahmad MA, and Srivastava J. 2008. Expert
Identification in Social Networks an Ant Colony Optimization Approach
. University of Minnesota
[30] Bolstorff P, and Rosenbaum R. 2007. Supply Chain Excellence - A Handbook for
Dramatic Improvement Using the SCOR Model
. Amacom.
ISSN 2086-5953 [This page is intentionally left blank]
109 ISSN 2086-5953
ARTIFICIAL NEURAL NETWORK TO MEASURE PHOSPHORUS AMOUNT IN AGRICULTURAL SOILS
Feriyonika
1
, Nasih Widya Yuwono
2 1
Graduate School of Automation and Control, National Taiwan University of Science Technology
2
Soil Science Department, Faculty of Agriculture, Gadjah Mada University Email: feriyonikayahoo.com
1
, nasihugm.ac.id
2
ABSTRACT
This research investigates alternative way to measure one of important elements, phosphorus, in
agricultural soils. In many agricultural lands, the farmers‘ behavior is just using their experience and
invalid information when take fertilizer in their soil. This condition not only scarifies the soil fertility
and harm microorganism but also waste the
farmer‘s money. This study develops equipment using camera, computer and Artificial Neural
Network ANN as algorithm. The result show that proposed method can do measure the phosphorus
amount well. Keywords: ANN, agriculture, soils, camera,
phosphorus.
1 INTRODUCTION
The Soil holds important role for the plant growth. It has main function as place to growth,
providing water and providing air to root breath. There are three main factors deciding the quality of
fertility in agricultural soil. They are Biology, physic and chemistry factors. In real world, they
influence one to another and must be equilibrium [1]. If the plants can get the nutrition directly from
the soil, they will grow well and result good crop.
Today, the nutrition for the plants provided by the agricultural soil is not enough. Since the
industrial Green Revolution of the 20th century, chemically synthesized inorganic fertilizers were
widely used in agricultural plant. It replaces organic fertilizer due to fulfill plant nutrition fast. Lately, it
has been realized that the utilization of the chemically synthesized inorganic fertilizers causes
decreasing of microorganisms in the soil and affect the decreasing of the soil fertility.
In Indonesia, a country with wide agriculture land, the utilization of the chemically synthesized
inorganic fertilizers has started since early 70-es. About twenty years later, early 90s, the crop has not
satisfied anymore.
The condition of soil is very important information for the farmer before they cultivate the
plant. The information will be useful to decide what the elements being needed are and how many is.
This procedure is not work well due to difficulties to conduct it. The farmers, who leave far from the
center of research or university, do the fertilizing based on experience and lack of information. This
condition causes disadvantage for the farmers; wasting fertilizer and money.
In this research, we use artificial neural network ANN to predict one of important
elements, phosphorus P, deciding the quality of soil‘s fertility. There have been researches in
agricultural field
involving soft
computing intelligence techniques. He et al. [2] use ANN to
estimate monthly total nitrogen concentration in streams. They investigate the relationship among
the land use, fertilizer, and hydrometerological conditions in 59 river basins over Japan and then
applied to estimate the monthly river total nitrogen concentration TNC. Anctill et al. [3] use ANN to
simulate daily nitrate-nitrogen and suspended sediment fluxes from a small agricultural catchment
Melarches, 70 km east of Paris, France. They investigate the soil erosion and biochemical
applications associated to human activities such as intensive agriculture. Lek et al. [4] used ANN to
predict phosphorus concentration and phosphorus load
from watershed
characteristics. They
investigate relationship between land use and the concentration and export of total phosphorus and of
orthophosphate by watersheds. Salehi et al. [5] used ANN to predict of annual nitrate-N N0
3
-N losses into drain flow at Eugene F. Whelan Experimental
Farm Agriculture Canada, Woodslee, Ontario, Canada. They used daily measurements of nitrate-
N taken from eight different soil conservation treatments during 1992-1994.
In this research, we use ANN to predict phosphorus amount in agricultural soils. The result
of this research is aim to make portable instrument included ANN inside which can be used in many
place easily. The instrument is helpful tool to give farmer information about soil condition. The
information is very important because farmer can give appropriate treatment to their agricultural land.
ISSN 2086-5953
2 MODEL, ANALYSIS, DESIGN,
AND IMPLEMENTATION
2.1 Proposed method