Internal Supply Chain Risk Management Using Failure Mode and Effect Analysis (FMEA) and Value at Risk (VaR) (Case Study in PT Agro Muda Berkarya)

—In the agricultural industry, the risk variability in the internal supply chain is very high which is influenced by various factors, namely humans, weather, nutrition, pests, and others. These risks must be mapped so that companies in the agricultural industry can avoid failed harvest, increase their productivity, and has the capability in producing quality crops. The high possibility of risk in the agricultural industry, especially in Indonesia, reducing the possibility of investment by private or foreign companies, this can be corrected by having a clear risk assessment in any agricultural commodity investment plans. The investment plan assessment is carried out in order to determine the probability of possible losses and profits. PT. Agro Muda Berkarya is a company engaged in the agricultural sector, producing and trading agricultural commodities on a national and international scale. This company, which is located in Bogor, generates its operational activities with an initial focus on food crop cultivation, PT. Agro Muda Berkarya started production with an initial capital of 5000m2 and produced ginger as main crop with luffa as the intercopping crop. In carrying out the planting process, PT Agro Muda Berkarya wants to know the risks that can arise and mitigation action throughout the internal supply chain and probability of return in the business investment. This business investment focuses on luffa and ginger plants. In this final project, identification and risk profiling was carried out using Failure Mode & Effect Analysis (FMEA) and Value at Risk (VaR) with the Monte Carlo Simulation methodology to determine the probability of losses and profits on investing in ginger and oyong as intercopping plant. Last, Return on Investment (ROI) were calculated in order to understand company business performace.


I. INTRODUCTION
NDONESIA has become one of the biggest agricultural country in the world. Most of the population works in the agricultural industry. Indonesia also has good natural resources that could be used to become th In Q3 2019, the agricultural sector contributes Gross Domestic Product (GDP) by as much as 13,45% country's priority sector to be developed. Therefore, based on this contribution, the agricultural sector becomes one of the main sectors in driving the economic development of Indonesians. Therefore, based on this contribution, the agricultural sector becomes one of the main sectors in driving the economic development of Indonesians. According to Global Business Guide, Indonesia is the world's largest producer of palm oil as well as coffee, rubber, cocoa, sugar, tropical fruits and spices. Nevertheless, among all favorite commodities, Indonesia still imports spices when in fact Indonesia has suitable land for producing it. This phenomenon occurs due to higher local spices compare to other South Asia spices. PT. Agro Muda Berkarya believe, current condition can be solved through investment in producing spices so that there is no supply and demand gap. PT. Agro Muda Berkarya is an agricultural company that engages in production and trade. Spices has become one of the main planting commodities due to the high demand and lack of parties that process spices on a large scale. PT. Agro Muda Berkarya also produces horticultural crops such as luffa (oyong) to support land productivity as well as the risk on cultivating spices plants. According to the data from the Ministry of Agriculture 2015 -2019, ginger has various levels of productivity, therefore creating variation that could lead to several risks. The probability of risk located in all internal supply chain PT Agro Muda Berkarya that refers to the chain of activities within a company that usually consist of production, distribution, purchasing, and sales. The internal company supply chain must be successful in order create  integration working environment and efficient system of work. The internal supply chain plays a big role in determining final product is ready and good to appear in the market and the demand can be fulfilled on time. Those activities in internal supply chain PT Agro Muda Berkarya are categorized in a form of business process such as precultivation, soil cultivation, cultivation process, corps maintenance, harvest and post harvest. Risk in every internal supply chain will directly affect the crops productivity. Those risks could be mapped by using other horticulture plants such as luffa (oyong) in order to reduce the risk of long-term ginger planting from the start of the planting until harvesting which takes approximately eleven months. although luffa (oyong) is intercropped to reduce the risk of ginger plants, there are risks that could have an impact to both crops productivity. The productivity influence by several factors, those factors are the variation of prices in the market, weather condition that directly affect plants nutrition, unrecognized pests and human error that leads to luffa and ginger productivity. Those risks are assessed by using various tools. In this final paper, the author considers to use Failure Mode & Effect Analysis (FMEA) because the method are able to identify and prioritized potential failure modes in a system, process, service or product [1]. The risk that already identified by using Failure Mode & Effect Analysis (FMEA) then perfected by implementing Value at Risk method using Monte Carlo simulation to understand the possibility of revenues and Return on Investment (ROI) on ginger commodity and inter-cropping vegetables.

A. Problem Identification Phase
The first step is to identify the problem, Author realize there is a risk in agriculture business that effect crops productivity which indicates loss of investment and several problems are felt by PT. Agro Muda Berkarya. By understanding the problem that occur in agriculture business, author can determine the purpose of the research which to assess risk in PT. Agro Muda Berkarya and create the investment feasibility. This process also includes benefits formulation, limitation and assumption in solving the problem. The next step is to to learn and understand the problem that want to be solve by looking for reference and methodology to conduct the research. Literature study consists of agriculture in Indonesia, Agriculture contribution to Indonesia economy, investment in agriculture, risk management in agriculture, risk management theory, Failure Mode and Effect Analysis (FMEA) for the basis of internal supply chain risk assessment, risk mitigation action theory for further recommendation, fishbone diagram theory to understand the causes of problems, pareto analysis for determining most of causes that influence problems effect, Value at Risk (VaR) to understand the basis financial risk assessment, Monte Carlo simulation to provide the possibility of revenues, and Return on Investment to understand how good is the business in terms of financial perspective (ROI). All the literature studies is carried out according to journals, books and trusted websites.
Next, field study conducted to understand PT. Agro Muda Berkarya business model include all the activities in company internal supply chain, reason in choosing a particular crop, cultivation details, production rate and      commodities price in market. Risk factors also validated by the company and farmers before risk mitigation recommendation given by the author. The data received by interviewing PT. Agro Muda Berkarya operational director and farmers. After conducting literature study, the methodology that author used has been determine and from the field study, author will understand the business model and internal supply chain of PT Agro Muda Berkarya. In this process, the alignment of theory and model to the problem already understood in order to create beneficial impact to the company.

B. Risk Assessment Phase
The first process in creating risk profile for PT. Agro Muda Berkarya is to list all the activities in internal supply chain. The activities include pre-planting spray, loosening ground, applying fertilizer, mulching process, create planting point, laying seeds, watering process, stake installation, maintenance fertilization, pest control, winding shoots and stems, picking the crop, crop sorting, and packing. After knowing each of activities, list all the potential risk that might occur, effect and causes. Before assessing the severity, occurance, detection in each of the activities, in this section need to find basis score and scale of each indicators from trusted journal. Figure 1 is the FMEA framework that author used in the final project research The next step is determining the severty, occurance and detection need to be done in order to know which effect is having the highest severity, probability of occurance and detection in PT Agro Muda Berkarya. The process of determining this process is accompanied by company operational director. After defining each of activities indictor, calculate the Risk Priority Number (RPN). The result of RPN calculation indicates which activity is the most dangerous in company internal supply chain.
After determining the RPN of each risk, start to define risk factor priority using pareto analysis. Pareto analysis is conducted in order to find out 20% of the causes effect 80% of problems. By using this methodology, PT Agro Muda Berkarya will know which causes need to be done immidietly. This process carried out by several steps, the first step is to convert each Risk Priority Number (RPN) into percetange, the second step is to create the cumulative percentage in each RPN, the third step is to create the pareto chart in order to have a understandable visualization 20% of causes effect 80% of problems.
Last, Risk mitigation consists of several action, which are avoiding risk, mitigating risk, transferring risk, and accepting the risk. In this process, all the effect of risk in PT. Agro Muda Berkarya internal supply chain will be determined by the form of risk mitigation.

C. Investment Assessment Phase
The first thing to do in making a model is to determine the variables that indicate revenue, which are plant productivity and crop market prices. Then, collect all the data needed. Data that has been collected especially luffa productivity data is validated using slovin test in order to know how much data will be representative.
After the data collection is done and representative and in each variable column is complete, determine the distribution of the data before simulating the model. Next, run the model to know the probability of revenue in this specific agriculture crops investment. Last, conduct a verification and validation. Verification model carried out by checking model design and program in order to check the model has been built according to the requirements or not. The objective of verification is to ensure the quality of the model, it involves activities such as inspection, walkthroughs and reviews. Meanwhile, validation is the process to ensure the model is representative to the real world by the help of expert judgement. According to Journal of Modern Applied Statistical Methods, 5000 replications are sufficient enough to produce stable results [2]. Therefore, author will have 5000 replications in order to create a stable result.
After generating monte carlo simulation, Return on Investment is carried out to understand how good the investment will be, this phase is conducted after all the possibility of revenue has occur from the simulation. Make sure the initial investment value is already determined before conducting the Return on Investment (ROI) calculation. Then calculate the present value of potential revenue in the future before compared to the initial investment in order to create a valid estimation considering there is a inflation in the coming year. Last calculate the Failure might cause death of the worker or other entity that invlove in the activity and/or system will breakdown without any warning 9 Very dangerous Failure might cause a serious injury and/or system breakdown with prior warning and need service after the event 8 7 Dangerous Failure might cause a minor to medium injury and might cause dissatification of stakeholders, event that occur required a major repair or re-work 6 5 Medium Danger Failure might cause a minor injury with small dissatisdaction from the stakeholders and/or major system problem 4 Low moderate danger Failure could cause a very minor injury but annoys stakeholder, easy repair or re-work, can be overcome with minor modification to the system or process 3 2

Small danger
Failure might not cause a major injury, stakeholder might not aware of the failure. little or no effect to the system 1 Zero danger Failure causes no trouble and injury to all stakeholder and has no impact to thr system The design control has a very low chance to detect a potential cause of failure 8 Remote Remote chance the design control will detect a subsequent failure mode 9 Very Remote Very remote chance the design control will detect a subsequent failure mode 10 Uncertain Design control does not detect any potential cause of failure or sabsequent failure mode ROI by deviding present value revenue with company initial investment in a specific period of time.

D. Risk and Business Investment Analysis Phase
In this process, the risk mitigation action that already determined before will be analyze in order to give a comprehensive understanding about reason in choosing a particular risk mitigation action. Furthermore, the Return on Investment will also be elaborated in this process in order to give a clear understanding on how good is the investment by looking the Return on Investment.

E. Conclusion and Suggestions
This phase will explain the conclusion based on the risk assessment analysis and investment analysis and based on the objectives of this research. Suggestions will be given for future research of similar topics.

A. PT Agro Muda Berkarya Company Profile
PT. Agro Muda Berkarya is one of the companies engaged in the agricultural sector, carrying out production activities and trading of agricultural commodities on a national and international scale. PT. Agro Muda Berkarya was established in 2020, producing with an initial capital of 5000 m2 and producing products of vegetables and spices namely: Luffa and Ginger. To run the 5000 m2, the total human resource that allocated in the farm is around 5 farmers, consist of one head farmer and four-day worker farmer, which only work when there are task that require couple of workers.
PT Agro Muda Berkarya implement intercropping cultivation system in order to achieve highly productivity investment and also several other benefits. Intercropping is one of the way to increase diversity in farming ecosystem by cultivating two or more crops simultaneously in one specific land for one season of cultivation.

B. Identification of Risk
The first step is to list all the internal supply chain activities inside PT Agro Muda Berkarya. From precultivation, soil cultivation, cultivation, maintenance, harvest until post harvest. To understand PT Agro Muda Berkarya internal supply chain, below is presented the activity flow chart in Figure 2.
After knowing all the activities occur, determine the possible risk that might occur in each of activities namely risk description. Table 1 presented the list of all activities in detail with the risk description.

C. Identification of Risk Causes and Impacts
Following the risk identification, the next step is to determine the causes and consequences of each risk. Causes and impacts that already determined will be consider as one of the properties in identifying risk priority number (RPN). Using this approach will enables the identification of root causes from several factors. Below presented the example of fishbone that is used to identify the causes of risk of "unsafe act by the farmers while using machine" Based on the fishbone diagram in Figure 3, Author capable of defining all causes that might occur in the internal supply chain. In addition, risk impacts also identified in each of the activities in order to easily define the severity scale of the potential risk. In Table 2 will show the example of detail of risk causes and impacts.
Above information has been identified by the Author, each of potential risk has their own impacts and causes. The amount of fertilizer is less or more than the ideal fertilizer needed on the soil The soil lack of nutrition 4 2 2 16 Based on above causes, the most causes is from unclear information in Standard Operating Procedure (SOP).

D. Identification of Risk Priority Number (RPN)
Knowing only potential risks, causes and impacts are not enough, it is very important to also understand which risk that might possess the most severe threats to the company. The Risk Prioriy Number (RPN) consist of severity, probability of occurance and detectability, all the three components are multiplied in order to know the final RPN number. Table 3, Table 4, Table 5 presented the criteria that is used in this final undergraduate thesis. Table 3 presented the severity scale that is used in PT Agro Muda Berkarya Internal Supply Chain Risk Assessment. Table 4 presented the occurance scale that is used in PT Agro Muda Berkarya Internal Supply Chain Risk Assessment. Table 5 presented the detection scale that is used in PT Agro Muda Berkarya Internal Supply Chain Risk Assessment. After defining all the Risk Priority Number (RPN) criteria, RPN each of activites are determined. In a particular condition, for severity that is above nine will directly become priority to handle. The calculation of RPN is carried out by multiplying the severity, probability of occurance and probability of detection. The RPN will indicates how important the risk and the impacts are. Table 6 will be presented the example of table that consist of risk description, impacts, severity, occurance, detectability value in each impact and final Risk Priority Number (RPN).
Each of risk description and impact has their own risk priority number, the above Risk Priority Number (RPN) only cover several risk descriptions from several internal supply chain activities. The RPN calculation need to cover all the risk description, all risk descriptions are elaborated in Table 1.

E. Pareto Analysis
Implementing analysis using Pareto, allow PT Agro Muda Berkarya to understand 80 percent of problems come from which 20 percent of causes. This analysis provides information that can help PT Agro Muda Berkarya to focus on particular problems. There are steps that Author conducted to carried out pareto analysis, first author list all the Risk Priority Number (RPN) from the highest to the lowest each of risk description and impacts. After all the RPN value already sorted, devide each of RPN value with the total of the RPN value to know each of risk contribution in percetange form. After that is to calculate the cumulative percentage. Last, create the Pareto chart using the RPN value and percentage to understand the Pareto visualization.
After following the steps in Figure 4, Author is capable to create Pareto Chart based on the highest to lowest RPN.
Based on the Pareto chart in Figure 4, the 80% point is in somewhere around EF10. Therefore, the 80% problems come from risk code EF1 until EF10. Those risks need to be priority in order to reduce risk in PT Agro Muda Berkarya internal supply chain. In Table 7 presented of risk and impact that is considered to be the priority.
Based on Table 7 risk priority summary, Author identified 10 risk that need to be mitigate, this does not mean that other risks are ignored and mitigation actions are not taken, but this is a priority that has a very large impact.

F. Risk Mapping and Mitigation Action
Determination of risk level is based on risk appetite by PT Agro Muda Berkarya by considering Risk Priority Number (RPN). Risk is divided into four levels, namely extreme risk, high risk, medium risk, and low risk. The level of risk acceptance is obtained from the results of company internal discussion. The results of the recap of the discussion regarding risk appetite produces a risk map in Table 8.
From the results of the calculation of the Risk Priority Number (RPN) and the mapping of risk appetite, Table 9 shows the results of the categorization of risk levels on PT Agro Muda Berkarya internal supply chain activities.
Risk level identification example in Table 9 are plot by the risk appetite that already determine in the Table 8, with the base of RPN. Each risk level has their own proportion, Figure 5 present a chart are the risk level proportion visualization for overall risk that exist in PT Agro Muda Berkarya.
According to the chart, there are only 5% extreme risk on PT Agro Muda Berkarya internal supply chain risk potential, and the biggest proportion is on medium risk with the amount of percetange is around 74%. Nevertheless, all the risk with all risk level need to be tackle in order to reduce and eliminate risk in the internal supply chain.
To be able to give a clear and comprehensive understanding to PT Agro Muda Berkarya, will be presented risk mitigation action that associated with the potential risk in each activity, causes, impacts and Risk Priority Number (RPN). Risk mitigation action can be defined into eliminate risk, reducing risk, transferring risk and accepting the risk. Figure 6 shows the proposed risk mitigation action, the character of action which are preventive means eliminate or reduce potential action that leads to risk and curative which means treat the risk that happened and action description. Based on Figure 6, there are no risk that are accepted because low risk did not identified in PT Agro Muda Berkarya internal supply chain.

G. Cultivation Investment
PT Agro Muda Berkarya has its own standards regarding cultivation. After the cultivation details are determined, investment regarding the luffa and ginger crop for 11 months must be presented because the data is needed to calculate Return on Investment (ROI) in the next chapter. Overall, there are four category investments, below is the summary of investment for ginger and luffa crop cultivation. Table 10 indicates the total investment for luffa and ginger cultivation for area 5000 square meters in Bogor, Jawa Barat. The total investment value will be use in the calculation of return on investment.

H. Monte Carlo Simulation
In Monte Carlo simulation that conducted by Author, will be presented the probability of revenue for the investment of ginger and luffa as the intercopping vegetables. There are several variable that is consider in creating the monte carlo simulation, which are luffa productivity, ginger productivity, luffa prices and ginger pieces from November 2020 until March 2021. To validate the amount of relevant data to be taken, a slovin test was conducted with the aim of knowing how many samples needed to be taken from the PT Agro Muda Berkarya farm, so that it will be representative. Below is the formula of the slovin test.

= +
(1) Where: = Number of samples = Total population e = Error tolerance Based on the slovin test, author determine the margin of error 0.05 because Author has the confidence level 95% and accurate enough to be stated representative. The population size is generated from the amount of PT Agro Muda Berkarya farm total trees and it is around 4009. Based on the calculation, Author need to gather 364 data regarding luffa crop in PT Agro Muda Berkarya farm.
After gathering all the data from crop productivity and market prices and ensuring the data representative, Author determine the distribution of the data before simulating the model. Next, run the model to know the probability of revenue in this specific agriculture crops investment. Last, conduct a verification and validation. According to Journal of Modern Applied Statistical Methods, 5000 replications are sufficient enough to produce stable results [2]. Therefore, author will have 5000 replications in order to create a stable result. Below is the simulation Monte Carlo result for 10 replications as the example.
The data in Table 11 that is presented above gathered from PT Agro Muda Berkarya farm on ginger and luffa commodities and for luffa and ginger price generated from market prices, all between November 2020 until March 2021.
Based on the Monte Carlo simulation, the probability of revenue can be obtained. The revenue is carried out from months of business investment focus at ginger commodity and luffa as the intercopping plant. From the simulation it can be conclude that the probability average revenue from the 5000 replications is Rp 283.676.545.

I. Return on Investment (ROI) and Value at Risk (VaR) Analysis
Author calculates the probability of return on investment from the cultivation of ginger and luffa as the intercopping crop. The calculation is carried out by dividing the present value of revenue to the initial investment that already inform in the Table 10. To calculate the ROI, Author adjusted the value of revenue to the future due the inflation. According to the central bank, growth forecast for 2021 interest rate is at might be 5.1%, therefore in the calculation of present value of revenue, Author use 5.1% as the interest rate for eleven months of investment. Table 12 presented the Return Based on the ROI calculation for 5000 replications, the average ROI is 28%. This number indicates how good the investment for almost one year in ginger and luffa commodities. There is 93% return on investment which bigger than 0% or positive, it indicates good opportunity to invest. There is number of risks for the investment for around 7%, but it can be tackle through segmented industry market and operational excellence at the farm to produce a quality and high productivity crops.
Furthrmore, Author need to calculate the ratio of potential lossess in PT Agro Muda Berkarya. Figure 7 is presented the ROI data distribution chart.
To interprate data using Value at Risk concept, Author calculate at 5% probability what is the possible ROI rate. With a 95% confidence level the worst investment for ginger and luffa as the intercopping crop loss will not exceed -3%.

I. CONCLUSIONS
Risk identification in PT Agro Muda Berkarya internal supply chain operational activities is carried out by identifying potential failures at each. Based on the results of the risk analysis carried out, Author obtained 17 types of risks, 17 types of impacts and 26 kind of causes in PT Agro Muda Berkarya internal supply chain activities.
Risk assessment is conducted using Failure Mode and Effect Analysis (FMEA) throughout PT Agro Muda Berkarya internal supply chain. Based on the risk assessment, the biggest RPN is on weather uncertainty because the cause is coming from external factor which is climate. Lowest RPN goes to the amount of fertilizer is less or more than the ideal fertilizer needed on the soil because the probability of occurance is small due to the exact amount of fertilizer given to the land and the probability of detection is also small because due to fertilizing scheduled activity. According to the pareto analysis, there are 10 risk that need to be priority from risk impact code EF1 until EF10 which has RPN value from 384 until 42.
The risk value is then categorized into risk levels based on the predetermined risk appetite. The risk appetite consist of extreme risk, high risk, medium risk, and low risk. The results of the risk mapping there is only 5% extreme, 21% high risk and the biggest proportion is on medium risk with the amount of percetange is around 74%. Other than that, the risk mitigation action is presented. In PT Agro Muda Berkarya case, there is no risk accepted because in risk assessment there is no low risk category. Above, there are 15 risk that will be eliminate, 3 that will be reduce and 1 that will be transfer.
Conducting Monte Carlo simulation will require several variables. There are investment value which is Rp128.520.000, Ginger and luffa productivity, and ginger and luffa market prices. All the data that Author acquire will be require for fitting distribution and then run the result using Microsoft Excel to determine the probability of revenue from this investment. The average of probability of revenue is Rp 283.676.545.
Return on Investment (ROI) calculation is conducted as the parameter of business investment in PT Agro Muda Berkarya. Author adjusted the value of initial investment to the future due the inflation. Author use 5.1% as the interest rate according to the central bank, growth forecast for 2021. The average ROI value is 28%, it can be conclude that the ROI is on the positive value. There are 93% ROI value that above 0% return and 7% ROI value is below 0%. Author believe it can be tackle by securing market such as luffa and ginger processing industry and through process excellence at the farm in order to create higher productivity. Author also calculate the ratio of potential lossess using Value at Risk (VAR) method, with a 95% confidence level the worst investment for ginger and luffa as the intercopping crop loss will not exceed -3%.