Proposed Improvement of Pertalite Fuel Outbound Logistics Performance to Achieve Pertamina Patra Niaga MS2 Compliance (A Case Study of Bandung City)

: PT. Pertamina Patra Niaga was founded in 2004, is a sub-holding of PT. Pertamina which handles Pertamina's downstream oil and gas processes, especially fuel trading and handling, as well as fleet and depot management. This study took Fuel Terminal Bandung Group - Ujung Berung as a research location which encounter issue related to the not achieved of the MS2 Compliance target in delivery time delay related to the external logistics performance. This research focuses on the delivery of Pertalite fuel in the city of Bandung in the third quarter of 2022. This study aims to identify the factor that affect the value of MS2 Compliance in Fuel Terminal Bandung Group-Ujung Berung and initiate the improvement initiatives to improve outbound logistic performance for Fuel Terminal Bandung Group - Ujung Berung. The conceptual framework used in this research describes the supply chain design and focuses on outbound logistics in terms of moving the final product to the retailers to achieve responsive supply chain strategy is improving the outbound logistic performance. The equipment that used to carry the delivery are tank trucks. The method used in this study uses a quantitative approach with heuristic methods using a sweep algorithm. Quantitative data were obtained from company data, observations at the company and interviews with departments that related to the company's logistics. Secondary data analysis using literature and company data. The analysis was carried out by finding the root of the problem with a fishbone diagram and analysing quantitative data with a sweep algorithm. The analysis is carried out by taking days that have low MS2 Compliance in Quarter III 2022 in one week, where one day has three delivery batch. From the results of the analysis, it was found that there were four main causes, namely method, environment, human, and equipment factors. Then examine by the sweep algorithm analysis which results in more effective route selection and maximizing the use of tank trucks by calculating the travel time for each shipment by considering the allowance time for each shipment.

compliance that are method, environment, equipment and manpower.The main focus of this research is to focus on the method and environment factors.For the method the author made efforts to solved the root cause with heuristic method namely sweep algorithm in vehicle routing problem.While the equipment root cause link with the method causes to maximize the tank truck utilization to minimize the effect of root cause from equipment factor.The manpower not discussed further because it is not related with a field of study.

REVIEW OF LITERATURE Supply Chain Design Strategy
The purpose of supply chain strategy is to balance the company's competitive advantage, namely between responsiveness and efficiency [2].From study the two case studies that already explain above concluded that primary tools to guide in the supply chain design process are enterprise databases, mathematical programming models, simulations, and geographical information systems.Due to current business conditions, the goal of supply chain to support the market strategy, and its design of supply chain must align accordingly [3].The design based on the target of the supply chain design strategy itself which reflect the design contents such supplier requirements and inventory targets that shown on the figure below [3].The author use the framework as same context for the comparison of efficient and responsive supply chains [2].

Table 1. Comparison of Efficient and Responsive Supply Chains
(Source: Chopra & Meindl, 2016) It can be seen on the table 1 there are comparison of efficient and responsive supply chain.The efficient supply chain primary goal is to give efforts to fulfill consumer demand at the lowest price by minimizing total costs and reducing costs (material cost, production cost, and inventory cost) along the supply chain [2].The responsive supply chains primary goal is the ability of supply chain in responding the rapid change in market demand [2].To achieve this responsive supply chain require a reliable distributor, reliable supplier that underlying speed and flexibility [2].Responsive supply chain is a supply chain's ability to do the following [2]: 1. Be able to respond to wide ranges of quantities demanded

Efficient Supply Chains Responsive Supply Chains
Primary Goal Supply demand at the lowest cost Respond quickly to demand

Product design strategy
Maximize performance while maintaining product costs down.

Pricing strategy
Reduced margins because customers are primarily motivated by price Higher margins since customers are not primarily motivated by price

Lowered costs due to high utilization
Retain its capacity flexible to protect against demand/supply volatility

Reduce inventory to save money
Manage buffer stock to address demand/supply volatility

Lead-time strategy
Reduce expenses, but not at the expense of others.

Deciding based on quality and cost
Decide depending on responsiveness, adaptability, reliability, and quality Fulfill a high service level The efficient supply chains effective when products operate with low profit margins, low product variation, and have a predictable demand forecast [4].While the responsive supply chains can be effective applicable when products have significant profit margins, a wide variety of products, and fluctuating demand, requiring swift adaptability to continually shifting consumer preferences [4].It can be concluded that supply chain design become the critical success factor for the whole business process.

Vehicle Routing Problem
In line with the logistic driver's the supply chain decision making framework the vehicle routing problem (VRP) as one of the tools to support from transportation decision [2].One of the most extensively researched subjects in the field of operations research is the "Vehicle Routing Problem" (VRP) [5].The aim of VRP is to account for the complexities of real-world situations, such as timedependent trip times (which reflect traffic congestion), time slots for pickup and delivery, and input information (such as demand information) that changes dynamically over time [5].The VRP has four main goals [2], which are as follows: 1. Reducing distance-related and vehicle-related fixed expenses connected with transportation.
2. Reducing the number of vehicles required to transport products.
3. Reducing costs brought on by agents' unsatisfactory service.Another reference said that the objective is to find and identify a group of delivery routes that satisfy certain conditions, limitation, constraints and have a minimum transportation cost [6].Many detailed methods, including guided local search and exact algorithms based on linear programming, have recently been utilized to address the VRP [6].
Some research has been done by researchers in the past discussing vehicle routing problem is that optimizing VRP using nearest neighbor method and saving matrix in PT XYZ [7].In the article, PT XYZ is a pharmaceutical company that provides distribution and logistics services in the health sector.The issue was difficulties when dealing with distribution process, specifically the frequent delay in delivering products to their customer due to the inaccuracy of route configuration and cause increase in transportation cost.The goal research is to determine distribution route and proper vehicle capacity at PT XYZ and to calculate the total cost of transportation by using nearest neighbor method and saving matrix [7].The result was by using nearest neighbor method it resulting six routes and using saving matrix, the company could save up to 27% and 38% respectively for each method [7].

Heuristic Method
An effective problem-solving strategy defined as a finite sequence of instructions is called an algorithm [5].Another way to look at it is that an algorithm is the combination of data structures and the algorithmic expression of a specific job [5].Heuristic algorithms are standards or computer processes designed to realize any purpose or select the most useful among several alternative actions to reach the goal [5].These algorithms have the ability to converge, but they can only guarantee a solution that is very close to the final one [5].These algorithms are referred to as improvable algorithms when they reach the best solution in the space of solutions [5].
In addition, heuristic methods have drawn a lot of attention from researchers working to solve complex VRPs such as neural networks, genetic algorithms, and evolutionary tactics are some of the heuristic techniques that have lately been put to use [6].Adding from newest research due of the size of real-world problems, heuristics and metaheuristics are frequently better suited for practical applications (e.g., a company may need to supply thousands of customers from dozens of depots with numerous vehicles and subject to a variety of constraints) [6].The angle (angular position) can be calculated by the formula below: The algorithm for sweep has been updated to become a linear algorithm [5].According to Figure 1, the nodes are split up between two depots at either end rather than a single central store.As a result, the originally mentioned sweep method cannot be used and must be modified.The x and y coordinates of the nodes, which are arranged in a graph, are assigned.

METHODOLOGY
In conducting this research the author uses the combination of data obtained by primary data and secondary data.This research collects data and analyses data using quantitative methods.Quantitative methods are defined a research method based on concrete data, research data in the form of numbers that measured using statistics as a calculation test tool, related to the problem being studied to produce a conclusion or in other words scientific method whose data is in the form of numbers or numbers that can be processed and analysed   The road condition in Bandung city is narrow, contoured up and down, and some road there are potholes so it need adjusting for big vehicle as tank truck.The example: From above case, total distance is 42.6 so the total time travel without allowance is 1) The major cause method have two causes that is ineffective selection of routes and hours of fuel delivery this happen due there has not any effective tool for vehicle routing.Other causes is difficulties in scheduling this happen due if there are outstanding demand from unpredicted demand so there is unplanned scheduling and routing.
2) The major cause environment have two causes that is unpredictable daily demand from consumer this happen due to retailers experienced unexpected their customer panic buying that affect unexpected increasing in daily demand during the day.Other causes is unpredictable traffic situation this happen due uncontrollable factor such as demonstration suddenly happen in Bandung City and flood in shipping routes.
3) The major cause equipment have one cause that is any presumptions where lack of fleets capacity because there are tank trucks that are maintenance and there are tank trucks that already over the lifespan >10 years.4) The major cause manpower have two cause that is indiscipline tank truck drivers and lack of tank trucks drivers but this major cause manpower cannot be discussed further because it is not included in the field of study.

Analysis of Heuristic Method (Sweep Algorithm) Table 2. The result of Analysis of Sweep Algorithm
It can be seen from the table 2, the sample was taken in Quarter III, namely July, August and September 2022, which is one full week from Monday to Sunday to see the delivery pattern.According to the analysis of sweep algorithm to answer the improvement initiatives to improve outbound logistic performance for Fuel Terminal Bandung Group -Ujung Berung define in each major cause derivative of analysis of root causes.To solved the method major root cause that is the implementation of new vehicle routing and shipping scheduling using heuristic method analysis using sweep algorithm to minimize or avoid the delivery delay.The result of heuristic method analysis is discovered that the total average of customer's location that have to delivered in shift is 32 locations with the average demand each delivery shift is 311238 L. Regarding the overall optimization of the use of tank trucks for each day, shift 3 is where it is most maximized because it lasts from 12 a.m. to 22 a.m., a period of time that is 10 hours longer than shifts 1 and 2. Additionally, the maximum tank truck use for delivery in shifts 1, 2, and 3 will vary depending on the proportion of traffic probability for each shift, which has variances.Additionally, shift 3 also experiences the lowest demand, whereas shift 1 experiences the greatest volume of shipments as a result of the urge for orders from a high number of clients.So that the average delivery batch in each day is 40 delivery batch.The delivery batch after maximization of tank truck utilization the average delivery batch in each day become 35 delivery batch, which will reduce the use of tank trucks.With the average travel time each day is 3.69 hours.Currently Fuel Terminal have 53 tank trucks for Pertalite products for the Bandung City for a day.Whereas, this should be sufficient to accommodate tank truck deliveries for Pertalite (from the results of mapping the distribution and capacity of tankers based on travel time using a sweep algorithm).Theoretically, it should not be late, but there are a number of factors that can decrease the effectiveness of scheduling as well as late routing.Because the sweep algorithm approach examines the largest angular angle, it can therefore minimize the tank trucks requirements or maximize tank trucks utility with more effective routing and route selection.Additionally, it can offer a summary and support for a more effective and reliable logistic network configuration with the sweep algorithm method.So that delays in delivery can be minimized.To solved the environment major root cause adding the allowance time to the time calculation, so the total time required for delivery is included the allowance time (unpredictable traffic).The allowance time consist of unloading time (Hours), set up time (Hours), %traffic probability (%) and filling time.this allowance time has been discussed according to company policy.The allowance of unloading time (Hours), set up time (Hours) and filling time have the same percentage of all delivery shift in each day.On the other hand, the allowance time of %traffic probability has different percentage because each shift has a different time which are:  Shift 1 (12pm-6am) = 10%  Shift 2 (6am -12am) = 40%  Shift 3 (12am -10pm) = 20% The differences of percentage of allowance time in traffic probability affecting the result to the travel time, delivery shift 2 and delivery shift 3 tend to have higher travel time rather than delivery shift 1. Therefore it also affect the maximization of tank truck utilizations in each delivery shift 2 and delivery shift 3.
To solved the equipment major root cause solved by maximizing the tank trucks utilization by calculating the total time travel, and maximizing the delivery using the same tank truck which has been used for the previous delivery on the same batch.The average delivery batch in each day is 40 delivery batch.The delivery batch after maximization of tank truck utilization the average delivery batch in each day become 35 delivery batch, which will reduce the use of tank trucks.Therefore, the lack of fleet capacity can be avoided due to the maximization of tank truck utilization.For manpower major root cause cannot be discussed further because it is not included in the field of study.

CONCLUSIONS
PT. Pertamina Patra Niaga is a sub-holding company of PT.Pertamina that handles downstream process of PT.Pertamina.The downstream process is including the distribution process of fuels which distributed to industries and gas stations throughout Indonesia.To support the business process, PT.Pertamina Patra Niaga accommodate Fuel Terminal that support the distribution process so that can reach all parts of the area covered by each Fuel Terminal.This research mainly focus only in Fuel Terminal Bandung Group Ujung Berung which cover the region near Bandung City, especially Bandung City.Along with the focus is the Pertalite fuel product, because Pertalite fuel product still have a high demand among other products.Due to the high demand of the Pertalite fuel product, Fuel Terminal Bandung Group Ujung Berung faced several issue directing the distribution process specifically in delay delivery time which a part of outbound logistic performance.This issue are reflected on the MS2 Compliance a part of key performance in PT.Pertamina Patra Niaga.Fuel Terminal Bandung Group Ujung Berung used Tank Trucks to distribute the Pertalite fuel product.Then it becomes crucial to improve the distribution process of Pertalite fuel product in Bandung City.
To solved this issue is analysing the delivery pattern of Pertalite fuel product in Bandung City that covered by Fuel Terminal Bandung Group Ujung Berung with heuristic method (sweep algorithm) as a part of vehicle routing problem.The sample is Quarter III 2022 namely July, August, and September 2022, and choose the day that showing decreasing number of MS2 Compliance with one week full to see the pattern in one week.The analysis result of this method is the capacity of Tank Trucks adequate along with

Figure 1 .
Figure 1.Illustration of Linear Sweep Algorithm (Source: Kumar & Jayachitra, 2016) using mathematical calculations or statistics (Sekaran & Bougie, 2017).Quantitative methods research that demands the use of numbers, starting from data collection, interpretation of the data, and the appearance of the results.The author collects quantitative data from recapitulation of MS2 Compliance on Quarter III 2022, total tank trucks, total customers, demand of each customers, type of tank truck capacity, coordinates via google maps, demand, mileage between customers, travel time via google maps and number of retailers as a customer served by Fuel Terminal Ujung Berung per region, and demand for each retailers on quarter III which are collected in PT Pertamina Patra Niaga specifically in Fuel Terminal Bandung Group Ujung Berung.The sample is MS2 recapitulation in QIII 2022 (July, August, September) which days in one week is taken where the dates used are the dates that represent the business issue which the MS2 value is low.The dates are:1) Monday, August 29 th 2022 2) Tuesday, July 5 th 2022 3) Wednesday, September 1 st 2022 4) Thursday, September 7 th 2022 5) Friday, August 12 th 2022 6) Saturday, July 2 nd 2022 7) Sunday, July 3 rd 2022 The purpose of take this sample is to analyse the pattern of customer's demand in one week and find the possible and effective routes on each day by considering the allowance time.From all data that are already gathered processed with quantitative methodology specifically in heuristic model that is sweep algorithm to determine the vehicle routing problem that resulting increase in MS2 Compliance value.This research use the sample of the date that showing low of MS2 Compliance which consist of one full week to see the Pertalite distribution pattern in the city of Bandung every day of the week.For this research, the author develop the conceptual framework (in figure2) as shown below refer to journal[8]: ISSN: 2581-8341 Volume 06 Issue 01 January 2023 DOI: 10.47191/ijcsrr/V6-i1-43, Impact Factor: 5.995 IJCSRR @ 2023 www.ijcsrr.org389 * Corresponding Author: Nastiti Liring Bestari Volume 06 Issue 01 January 2023 Available at: ijcsrr.orgPage No. 385-394

Figure 2 .o 3 .
Figure 2. Conceptual Framework to be used in this research

o
time The allowance time consist of unloading time (Hours), set up time (Hours), %traffic probability (%) and filling time.this allowance time has been discussed according to company policy.o Unloading Time (Hours) = time to move (transfer) Pertalite to the tank at the gas station.o Set Up Time (Hours) = time to prepare the vehicle to enter the gas station and clear the area.o Percentage (%) Traffic Probability Hours = time allowance for possible congestion in each delivery shift.Multiplied by the total travel time without allowance.Shift 1 (12pm-6am) = 10%  Shift 2 (6am -12am) = 40%  Shift 3 (12am -10pm) = 20%  Filling Time in Fuel Terminal consist of: o Administration o Waiting time into Filling Shed o Pre & Post Loading o Quality control From all section of filling time in Fuel Terminal the total filling time is 40 Minutes.The total allowance time calculated by adding filling time, unloading time (Hour), set up time (Hour), %traffic probability.The example:  Shift 1: o Filling Time = 40 minutes = 0.67 hour o Unloading Time (Hour) = 30 minutes = 0.5 hour o Set Up Time (Hour) = 15 minutes = 0.40 hour o % Traffic Probability = 10% x total travel time for each batch  Shift 2: o Filling Time = 40 minutes = 0.67 hour o Unloading Time (Hour) = 30 minutes = 0.5 hour o Set Up Time (Hour) = 15 minutes = 0.40 hour o % Traffic Probability = 40% x total travel time for each batch  Shift 3: o Filling Time = 40 minutes = 0.67 hour o Unloading Time (Hour) = 30 minutes = 0.5 hour o Set Up Time (Hour) = 15 minutes = 0.40 hour o % Traffic Probability = 55% x total travel time for each batch 2. Total Travel Time With Allowance Time Travel Without Allowance + Allowance time Example: From above case, Time Travel Without Allowance = total distance is 42.6 Km 40 Km/Hour = 1.07 Hour.o The allowance time for shift 1 = unloading time (0.5 Hour) + set up time (0.4 Hour) for each gas station + % traffic probability = 2.8 Hour The Total Travel Time With Allowance for one batch delivery is 3.87 Hour Each batch delivery calculated as above explanation, then categorized the Total Travel Time With Allowance which is less than 2 hours.To maximize the utilization on tank truck capacity in Fuel Terminal. Shift 1 (12pm-6am) = ≥ 2.5  6 Hours  Shift 2 (6am -12am) = ≥ 2.5  6 Hours  Shift 3 (12am -10pm) = ≥ 4 10 HoursRESULT AND DISCUSSION Analysis of Root CausesAccording to the analysis of root cause using Ishikawa fishbone diagram that has been done in IV.1, the factor that affect the value of MS2 Compliance in Fuel Terminal Bandung Group-Ujung Berung are divided into 4 major causes namely Method, Environment, Equipment, and Manpower.