Analysis of container loading-unloading performance at Teluk Lamong Terminal Surabaya

PT. Teluk Lamong Terminal (TTL) is a container terminal whose primary business is to serve the loading and unloading of containers. As a result, shipping by water is necessary and will become more so with each passing year. This was seen between 2017 and 2021, when the arrival of ships and the flow of containers at PT. Teluk Lamong Terminal increased. The goal of this study is to determine how strong the association between the factors investigated is the loading and unloading performance (container flow) at Teluk Tamong Terminal Port. This research employs both qualitative and quantitative approaches, as well as primary and secondary data sources. Using the regression technique, the container flow of Teluk Lamong Terminal in 2031 is 1,357,208 TEUs, with a growth rate of 4.49%, falling short of the UNCTAD criterion. The results revealed a substantial positive association between the variables Idle Time (X1), Not Operation Time (X2), ASC Utility (X3), and Container Flow (Y), with Fcount 276.730 > Ftable 19.2 and a significance of 0.044 0.05 accepting Ho and Ha. The Adjusted R Square score for the association between Idle Time (X1), Not Operation Time (X2), ASC Utility (X3), and Container Flow (Y) is 0.995, indicating a strong relationship.


Introduction
Indonesia is a maritime country with a large enough ocean so that sea lanes are a means of transportation and distribution between provinces in Indonesia.Under these conditions, the maritime sector is of great significance to the goods and passenger transportation system for both domestic and international shipping.As a maritime country, in supporting inter-island activities, it must have adequate infrastructure.One of the supports for inter-island activities is the availability of ports.
A port is a facility that functions as a place to switch modes of transportation and move goods, cargo and passengers in it [1].In loading and unloading services, obstacles often occur in operations.In addition, loading and unloading productivity is not optimal due to lack of maintenance tools so that it has a direct impact on the productivity of loading and unloading equipment.These obstacles can result in losses for the company, ineffective loading and unloading equipment.
Furthermore, in research conducted by [2] where in the conditions of loading and unloading equipment and adequate human resources, the BCH standard is 25 boxes but the realization only reaches an average of 18 boxes.Each type of tool has a capacity with advantages and disadvantages of each that can affect loading and unloading productivity.The results showed that there was a partially significant positive relationship between the variables Crane (X₁) and Berthing Time (Y), where tcount 4.839> ttable 2.03224 and significantly 0.000 <0.05, then Hₒ was rejected and Ha was accepted.Judging from the correlation coefficient between Productivity Crane (X₁) and Berthing Time (Y), 0.639 indicates a strong relationship because it is within the interval (0.600-0.799).

Loading unloading
Loading and unloading activities are activities of unloading goods from on the ship by using the crane and ship sling to the nearest land in edge of the ship which is commonly called the wharf, then from the wharf with using lorries, forklifts loaded and arranged into the nearest warehouse appointed by the harbor master.While loading activity is the opposite activity.Loading and unloading operations from/to ships.

Five whys analysis
Five why analysis is a simple root cause analysis tool and can be used to analyze system failures and can work well in identifying the causes and effects of an event [3]. 5 why analysis is used to dig deeper to the real root of the problem, the root cause can be known by asking "why" repeatedly until it reaches a point where the answer to the question has shown a root problem [4].

Multiple linear regression
The multiple linear regression method is a probabilistic model which states a linear relationship of more than one variable where one variable is considered to influence the other variables [5].The following is the formula for the multiple linear regression method: Which Y = Dependent Variable β0, …, βn = Regression Coefficient X1, …, Xn = Independent Variable

Classical assumption test
The classical assumption test is a statistical data test that is used as a prerequisite to provide certainty that the regression equation obtained has accuracy in estimation, is not biased and is consistent.
1) Normality test Normality test aims to determine whether the data distribution deviates or not from the normal distribution.Data that is good and feasible to prove the research models is data that has a normal distribution.

2) Heteroscedasticity test
A heteroscedasticity test was conducted to analyze whether a linear model has a different variance from the residual of one observation to another or not.Heteroscedasticity indications can be seen from the difference in the residual variance of each independent variable.To indicate heteroscedasticity, the Spearman test and scatter plot graphs were carried out [6].

3) Multicollinearity test
The multicollinearity test can be determined by carrying out the Variance Inflating Factor (VIF) test.VIF is an easy method for analyzing the multicollinearity of data.In the VIF test it can be observed when the t (tolerance) value is above 0.1 and the VIF value is below 10, it can be said that there is no multicollinearity between the variables being analyzed or vice versa [6].

4) Autocorrelation test
The autocorrelation test is analyzed to see whether there is a correlation between the error in the period (t) and the error in the previous period (t-) in a regression model or not.The autocorrelation test is carried out using the Durbin -Watson test or the DW test.

Loading-unloading performance
In this study the loading and unloading performance is used as the object of research to analyze the factors that affect loading and unloading performance.The data used in the analysis process are Idle Time, Not Operation Time, and the Automated Stacking Crane Utility.

Five whys analysis
Then a search for the root causes of the delay in the loading and unloading process and not achieving performance standards at Teluk Lamong Terminal was found by brainstorming and interviews, then processed using 5 Whys Analysis.

Classical assumption test
The classical assumption test is a statistical test that is used as a prerequisite to provide certainty that the regression equation obtained has accuracy in estimation, is not biased and is consistent.The Kolmogorov-Smirnov test above shows the significance of all variables around 0.200 where the value is above the limit_(standard error), 0.05 or 5%.This shows that the variable data tested is normally distributed and can be continued for other classical assumption tests.

2) Heteroscedasticity test
Spearman's rank test can be done using SPSS with regression of all independent variables on residual values.If there is a significant independent variable influence (Sig <0.05) then it can be said that there is a heteroscedasticity problem.To analyze whether there are signs of heteroscedasticity, the significance of each variable can be used by regressing the residual variance of each independent variable.If the table significance value of each independent variable is above 0.05, it can be said that there is no heteroscedasticity.If the table significance value of each independent variable is below 0.05, it can be said that there is an indication of heteroscedasticity.Based on the table above, the significance value of each independent variable shows a value above 0.05 (5%).This shows that there is no sign of heteroscedasticity among the independent variables to be analyzed, so that the data is homoscedastic.

3) Multicollinearity test
The multicollinearity test can be determined by using the Variance Inflating Factor (VIF) test.
The VIF test is an easy test method for analyzing data multicollinearity.To analyze indications of multicollinearity, t and VIF values can be observed if the t (tolerance) value is above 0.1 and the VIF value is below 10.It can be said that there is no multicollinearity between the variables being analyzed and vice versa.The multicollinearity test above shows that the VIF of each variable is above 10 and the tolerance value is above 0.1, so that each variable is independent of multicollinearity.

4) Autocorrelation test
The autocorrelation test was carried out to analyze whether there is a correlation between the tperiod error and the previous period error (t-) in the regression model or not.The autocorrelation test was carried out using a run test.The results of the run test are shown in the table below.

Multiple linear regression
Multiple linear regression tests were carried out on the variables Idle Time (X1), Not Operation Time (X2), and ASC Utility (X3) which can be seen in the table below.The linear equation obtained in equation 2 can be visualized through figures 1, figure 2, and figure 3, where the y-axis represents the sum of TEUs.And the conclusion wherein figures 2 and figure 3 exhibit a similar pattern, culminating in an increase at the end.

Conclusions
From the results of the analysis that has been carried out regarding loading and unloading performance, the following conclusions are obtained: • Factors that influence the failure to achieve loading and unloading performance standards are due to human, machine and environmental factors.• From the results of the Classical Assumption Test with 3 independent variables including IT, NOT, and ASC Utility, the results show that the data used is normally distributed, there are no indications of heteroscedasticity, multicollinearity, and autocorrelation.• In the Multiple Linear Regression results, the most influential variable on loading and unloading performance is the ASC Utility with a constant value of 25838.957.

Table 1 .
1) Normality test In the normality test the software is used to obtain significant values as follows.Result of normality test.

Table 2 .
Result of heteroscedasticity test.

Table 3 .
Result of multicollinearity test.

Table 4 .
Result of run test.Based on the table above, it can be concluded that there is no autocorrelation problem in the regression model.

Table 5 .
Result of multiple linear regression.Based on the results of the table above, the equation to form a linear regression model can be seen in the Unstandardized Coefficients table in column B. Based on the analysis results, a linear regression model can be formed as follows.