Data Visualization and Forecasting Domestic Component Level (TKDN) Indonesian Ministry of Industry Using Power Business Intelligence

. The government is optimizing the Domestic Component Level (TKDN) for strategic projects funded by the State. Domestic products that must be used must have a TKDN value of at least 25%. According to the Head of LKPP, the target for procurement of government goods/services for domestic products (PDN) in 2023 is 90%. Ironically, products that have TKDN certificates are not comparable to the needs of government agencies. This is of particular concern, whether the targets declared in 2023 will be met or not. Therefore, a touch of technology is needed to be able to analyze TKDN data quickly and precisely. Power Business Intelligent is one of the tools that can be used to analyze data easily understood. Data is taken on the official website of TKDN Kemenperin which has a Microsoft excel file extension. The indicators observed are the trend of TKDN certificates per year, the trend of certificates per month, the trend of goods categories, regional demographic trends in Indonesia and forecasting TKDN certificates for the next 5 years. The overall results of TKDN certified products in 2019-2022 were 33,554. The average number of TKDN products is 49.15% with the highest content of 99.99%. On average, the Ministry of Industry issues TKDN certificates every day as many as 43 products. The number of products certified TKDN with the LKPP e-catalog website is not comparable. The results of the analysis showed that the number of new TKDN certified products was 34.6% and 65.94% had not been certified.


Introduction
Currently, the government is optimizing the Domestic Component Level (TKDN) for strategic projects funded by the State.Based on the regulation of the Minister of Industry of the Republic of Indonesia number 16/M-IND/PER/2/2011 that the domestic component level (TKDN) is the amount of domestic components in goods, services and a combination of goods and services.TKDN of goods is calculated based on the comparison between the price of finished goods minus the price of foreign components against the price of finished goods.
The price of finished goods is the production cost incurred to produce goods [1].
Procurement of goods and services, users of domestic products are required to use domestic production if there are domestic products that have a sum of TKDN values and Company Benefit Weight (BMP) values of at least 40% [2].Domestic products that must be used as intended must have a TKDN value of at least 25% [3].The process of obtaining TKDN certification is that the industry can submit an independent calculation (self-assessment) regarding the TKDN value in its products.The results of the independent calculation are then verified by the Independent Verification Institutions appointed by the Ministry of Industry, namely PT Surveyor Indonesia and PT Sucofindo, then the TKDN certificate is issued by the Ministry of Industry [4] Procurement of goods is currently recommended to use electronic systems, one of which is the e-catalog issued by the Government Procurement Policy Institute (LKPP).In 2022, the realization of procurement for domestic products has reached 78%, both for providers and self-management.Therefore, according to the head of LKPP, the target for procurement of government goods/services for domestic products (PDN) in 2023 is 90% [5].The number of products that have TKDN certificates is not in line with the target target of procurement of government goods/services for domestic products (PDN) in 2023.This is of particular concern, whether the targets declared in 2023 will be met or not.Therefore, a touch of technology is needed to be able to measure or analyze TKDN data quickly and precisely.Power Business Intelligent is one of the tools used to analyze data with visualizations that are easy to understand and can also forecast the number of certified products in the next 5 years.So that with this data, stakeholders can see when the product is TKDN certified thoroughly.

Time Series Forecasting
Forecasting is one technique in data mining that is useful for making knowledge about data in order to predict future data by learning from previous datasets or data [6,7].Time Series is a set of observations where the variables used are measured in a sequence of time periods, such as yearly, monthly, quarterly, and so on.Time series is also a forecasting method based on the use of analysis of the pattern of relationships between variables to be estimated with time variables.The goal is to determine patterns in historical data series and extrapolate those patterns into the future [8].Forecasting methods are generally divided into two according to Eddy Herjanto (2004) [8], namely: first quantitative, and quantitative.There are four types of data patterns in time series forecasting, namely trend patterns, seasonal patterns, cyclical patterns and irregular patterns.so that if the time series data is symbolized by the variable Y, then Y is the multiplication between trend, cyclical, seasonal and irregular factors: Where T is trend, C is cyclic, S is seasonal, and I is irregular.Among the four components of the time series, the trend factor or future data trend is the most important factor.Thus, the purpose of periodic series analysis is related to efforts to find trend models and their usefulness in data prediction.
The least square method is an attempt to minimize the square result between the original data and the prediction data.In principle, a trend model is similar to a regression model that uses the least square method: Where: ¥ = Y is the result of the prediction, where Y itself is the original data of the time series.X = time-related code Because in a time series, the number X is zero, then the average is also equal to zero, so the formula [9]: (3) So that in the time series forecasting research using quantitative methods, namely forecasting is based on TKDN certificate data from previous years.

TKDN Kemenperin
TKDN (Domestic Component Level) is a limit or value that represents what is the level of domestic local content in a product of Goods / Services.In measuring TKDN for a product, there are 3 aspects that will be assessed, namely Material, Labor, General Service Costs (Overhead).First, materials are valued based on country of origin, meaning that they are traced, made, and produced.Second, the labor employed is assessed based on nationality.Third, Overhead (work tools/machines, and other costs related to the production of these Goods/Services).The main purpose of TKDN, theoretically, is to get rid of everything related to imports and intensify local businesses and even grow new businesses so that both materials and overhead can use local components and production costs are minimized [10].
Government policy in encouraging the use of domestic products has actually been regulated in such a way, so that industries in Indonesia have begun to improve themselves in maximizing the use of domestic products in all activities of their industrial sectors.We can see this in Law Number 3 of 2014 concerning Industry article 86 which reads The obligation of K/L/PD/BUMN/BUMD/BU Private and/or to cultivate resources controlled by the state using PDN in every procurement of goods/services [3].

Power Business Intelligence
Business Intelligence (BI) is a set of theories, methodologies, processes, architectures, and technologies that transform raw data into meaningful and useful information for business purposes [11].BI can handle large amounts of information to help identify and develop new opportunities.Taking advantage of new opportunities and implementing effective strategies can provide competitive market advantage and long-term stability [12].BI is also able to analyze, and present large amounts of data [13].In addition, business intelligence is an architecture and an integrated collection of operations and is also a decision support application and database that provides easy access for business communication to business data [14].The most fundamental objective is for the extraction and exploration of data from the Data Warehouse associated with the BI system, suitable for users and organizations to improve the quality of information, acquire the knowledge necessary for decision making [15].In accordance with the rapid development of technology, the field of BI science has also undergone several conceptual changes as shown below [16].

Fig. 1. Evolution of BI [17]
In this study, using the Power BI application to help people see the types of jobs that exist.
Power BI is an application developed by Microsoft that can be used for data analysis and reporting.Using this Power BI application can make it easier to analyze and forecast TKDN quickly and easily [12].
Power BI is a reporting and analysis software package released by Microsoft in 2015 [18].
Power BI is a collection of software services, apps, and connectors that work together to transform data from different sources into coherent, visually engaging, and interactive reports [7].Power BI consists of several elements that are integrated with each other, starting with the following three basics:  Power BI Report Server, to create reports on an on-premises server that can publish reports.

Method
This research generally has several stages in its implementation, including: literature study, problem analysis, data collection, data processing, data visualization, forecasing, analysis, and report results.Figure 3 shows the flow of research methods carried out in this study:  Starting from the TKDN certificate excel file data source downloaded from the official website of the Ministry of Industry, then saved to the MySQL server database and run to the web server from Power Business Intelligence.The indicators observed are the trend of TKDN certificates per year, the trend of certificates per month, the trend of TKDN based on the category of goods, the trend of TKDN based on regional demographics in Indonesia and forecasting the growth of TKDN certificates in the next 5 years.Some indicators are created using graphical data visualizations that are easy to analyze quickly.Data visualization is published using websites and mobile access with the aim of facilitating the process of data analysis by researchers and the wider community.

Results
The results of the ministry of industry's time series forecasting TKDN research using Power Business Intelligence which was carried out obtained interesting and very important data visualization results and discussions, including the following:

. TKDN Certificate Data Visualization
The number of certified products in 2019 was 3,207 (three thousand two hundred and seven), in 2020 there were 5,886 (five thousand eight hundred and eighty-eight) products, in 2021 it increased significantly by 15,075 (fifteen thousand seventy-five) products, and finally in 2022 it decreased with the number of TKDN certified products as many as 9,386 (Nine thousand three hundred and eighty-six).The advantages of Power BI are being able to display data visualizations with drill up and drill down.As in Figure 3, dril down can be done to see the results of TKDN data in certain quarters, monthly, and even daily.The following are the trends of TKDN certificates in the quarter of 2022.a large number of baling as many as 4,285 (four thousand two hundred and eighty five), and finally the least was in quarter 4 sebayak 803 (eight hundred and three) products.The small product numbers in Q4 can be analyzed by dril down to get a visualization of per-month data such as Figure 7.The TKDN group of goods is the most in Health Materials and Equipment as much as 18.18%, the second is Other Goods as much as 16.16%, the third is the Electrical Equipment group as much as 16%, the fourth is the Building / Construction Materials group 8.9%, the fifth is the Chemicals and Chemicals group as much as 8.26%, the sixth is the Metal and Metal Materials group is 7.63%, the seven groups of Oil and Gas Machinery and Equipment as much as 5.47%, the eight groups of Electronic Equipment as much as 4.46%, the nine groups of Agricultural Supporting Materials as much as 4.01%, the tenth group of Telecommunication Equipment as much as 3.28%, the eleventh group of Computers and Office Equipment as much as 1.81%, the twelfth group of Agricultural Machinery and Equipment as much as 1.57%, the thirteenth group of Clothing and Work Equipment as much as 1.56%, the fourteenth group of Sports and Education Equipment as much as 1.07%, fifteenth, namely the Transport Equipment group as much as 0.46%, sixteenth group of Defense Facilities as much as 0.45%, seventeenth namely in the Factory Machinery and Equipment group as much as 0.34%, eighteenth namely the Mining Machinery and Equipment group as much as 0.17%, and finally Heavy Equipment, Construction and Material Handling as much as 0.2%.The percentage of goods group Figure 9 corresponds to the number of products that have been carried out TKDN certification by the Indonesian Ministry of Industry in Figure 10 below.

Fig. 10. Number of TKDN by Item Group
The largest number of products in the group of goods is in Health Materials and Equipment of 6,099 (six thousand ninety-nine).While the smallest group of goods is Heavy Equipment, Construction and Material Handling as many as 67 products.Power BI can intelligently detect product locations based on region names around the world.It is proven in industrial address data, including industrial provinces, able to map areas as follows:.Demographics Figure 11 shows a data visualization using Map Location Power BI with a color ball visual.The larger the size of the sphere represents a large amount of data.Vice versa, if the size of the color ball is small then the data is also ensured to be small.More complete amounts of data and their percentages can be seen in Table 1 below.The most TKDN certified products by the Ministry of Industry are in the province of West Java by 29.36% with a total of 9,851 products (sembian thousand eight hundred and fifty one).While the second is in DKI Jakarta province by 18.86% with a total of 6,329 products (six thousand three hundred twenty-nine).The third largest rank is in East Java province at 17.46% with 5,859 products (five thousand eight hundred and fifty-nine).While the least number of TKDN products is 0.01% in Bengkulu province with 2 (two) products.

Forecasting
Through time series which is a collection of observations where TKDN variables are measured in the order of time periods, years, months, quarters, and so on.Time series is one of the features found in the Power BI application.Forecasting is based on the use of pattern analysis of the relationship between the variables to be estimated and the time variable.The goal is to determine patterns in historical data series and extrapolate those patterns to the next 5 years as shown in Figure 12.Based on Table 2 that in 2023 the number of TKDN certified products is 13,879 (thirteen thousand eight hundred and seventy-nine) with the highest predicted value of 20,838 (twenty thousand eight hundred thirty-eight) and the lowest of 6,921 (six thousand nine hundred twenty-one).Meanwhile, in 2024, the number of TKDN-certified products is 16,431 (sixteen thousand four hundred thirty-one) with the highest predicted value of 23,425 (twenty-three four hundred twenty-five) and the lowest of 9,437 (nine thousand four hundred thirty-seven).
In 2025, the number of TKDN certified products is 18,983 (eighteen thousand nine hundred and eighty-three) with the highest predicted value of 26,012 (twenty-six thousand twelve) and the lowest of 11,953 (eleven thousand nine hundred and fifty-three).In 2026, the number of TKDN certified products is 21,534 (twenty-one thousand five hundred thirty-four) with the highest predicted value of 28,600 (twenty-eight thousand six hundred) and the lowest of 14,468 (fourteen thousand four hundred sixty-eight).In 2027, the number of TKDN certified products is 24,086 (twenty-four thousand eighty-six) with the highest predicted value of 31,188 (thirty-one thousand serhundred eighty-eight) and the lowest of 16,983 (sixteen thousand nine hundred and eighty-three).

Fig 2 .
Fig 2. Three Elements Power BI All three elements (Power BI Desktop, service, and mobile app) are designed to enable you to create, share, and use business insights in the most effective way.Beyond these three, Power BI also displays two other elements:  Power BI Report Builder, to create a multi-page report to share in Power BI service.Power BI Report Server, to create reports on an on-premises server that can publish reports.

Fig. 3 .Fig. 4 .
Fig. 3. Research Method Flow In Fig 3 there are 6 steps to approach research with the following details: a. Literature Study: Research begins with an assessment process related to the research topic taken.In this study, the references used were obtained from journals that have a relationship with the classification of non-functional needs.b.Problem Analysis: The technique of problem analysis as a way to detail a depth of several elements becomes so important, in order to obtain information.The most basic problem is that the number of TKDN certified products of the Ministry of Industry has not been able to meet the needs of users.c.Data Collection: The data collection process uses the Ministry of Industry's TKDN datasheet.The data can be accessed on the website http://tkdn.kemenperin.go.id/rekap.php,d.Processing Data: Datasheet will be processed by cleaning data to remove inconsistent data and noise.e.Data Visualization: Data that has been consistent will be visualized using the Microsoft Power BI application program by displaying several dashboards that are needed for research.f.Forecasting: Forecasting is carried out using the Microsoft Power BI application program tool with a Time Series approach to quantitative data on the number of TKDN certificates in the previous year.So that the forecasting process of the number of TKDN certificates for the next 5 years can be found.g.Analysis: The analysis process is carried out based on the results of data visualization and TKDN forecasting by comparing the number of products displayed in the LKPP ecatalog.Website address: https://e-katalog.lkpp.go.id/ h.Results Report: The results of the analysis are made research reports ranging from literature studies, problem analysis, data collection, data processing, data visualization, forecasing, analysis, and result reports.In addition, the research results are published in national journals, so that stakeholders can see and study TKDN data clearly and easily.Research on Data Visualization and TKDN Forecasting of the Ministry of Industry Using Power Business Intelligence can be seen from the system architecture below: /doi.org/10.1051/e3sconf/20234480200303 (2023) 448

4. 1
Data Visualization TKDN certificate trend visualization presents a graph of the number of products that have been certified by the Ministry of Industry (Kemenperin) of the country of Indonesia from 2019 to 2022 as in Fig 5.

Fig. 6 .
Fig. 6.Quarter 2022 Data Visualization Quarter 1 (one) TKDN certified products were 1,355 (one thousand three hundred and fifty five), quarter 2 as many as 2,943 (two thousand nine hundred forty three), quarter 3 received

Fig. 7 .
Fig. 7. Data Visualization for Quarter 3 of 2022Based on Figure7which givesinformation that certified products are only carried out in October, so this greatly affects the decline in trend in quarter 3. Power BI is able to dril down October to display data visualizations per day as follows.

Fig. 8 .Fig. 9 .
Fig. 8. TKDN Certificate Data Visualization for October 2022 According to Figure 8 data, in October, the TKDN certification process lasted 10 days, namely on 2, 4, 5, 6, 7, 10, 11, 12, 13, and 14.The number of certified products fluctuates (not fixed or changing) with an average number of certified products of 29.There are several groups of goods classified by the Ministry of Industry that can be visualized in the form of pie charts as follows: /doi.org/10.1051/e3sconf/20234480200303 (2023) 448

Fig. 12 .
Fig. 12. Grafik Forecasting TKDN It is known on the forecasting chart using Power BI with forecast length parameters of 5 (five), Seasonality of 1 (one) and Convidence Interval of 95% to predict an increase in the number of TKDN certifications in the next 5 (five) years.Seasonality serves to determine the trend of performance of a historical TKDN data.While Convidence Interval is a parameter used to determine the accuracy of the Mean TKDN data.The results of forecasting or prediction there are 3 important data, namely (1) Forecast Value which is the reducing value, (2) Confidence Hight Bound which is the highest prediction value, (3) Confidence Low Bound which is the lowest predicted value.

Table 1 .
Number of TKDN Products at the Provincial Level