The Analysis of Road Condition on the Performance of Muara Teweh – Puruk Cahu Road Section

The existing condition of Muara Teweh-Puruk Cahu road section with ± 100 km long is mostly uphill and bend that currently is experiencing significant damage. This study aims to determine the type of damage in Muara Teweh-Puruk Cahu road section and to analyze the correlation between damage and vehicle speed on flat roads and hilly terrain. The study was conducted by collecting primary data in the form of data of road damage, vehicle speed, and road geometric. Data analysis of road damage is done by the BinaMarga method to obtain the type of damage. The correlation between road damage and vehicle speed is shown in a regression model. The results of the analysis show the vehicle speed will tend to decrease as the increasing number of road damage. The speed of light vehicles on roads with flat terrain shows a very strong correlation with the type of damage, such as cracks, grooves, and depression. The speed of light vehicle going uphill shows a very strong correlation with the type of damage, such as cracks, grooves, potholes, and depression. Meanwhile, when the vehicle goes downhill, it is seen a strong correlation with the types of damage such as cracks, grooves, and depression. The speed of heavy vehicles on roads with flat terrain shows a very strong correlation with the type of damage like cracks, grooves, potholes, and depression. The speed of heavy vehicle when go ing uphill shows a very strong correlation with the type of damage like cracks, grooves, potholes, and depression. Meanwhile, when the vehicle goes downhill, it is shown a very strong correlation with the types of damage like cracks, grooves, and depression. Keywords— Road Damage, Vehicle Speed, Regression Analysis.


INTRODUCTION
The condition that road infrastructure is burdened by traffic volume with heavy burden repeatedly will cause a decrease in the quality of the road. As an indicator, it can be seen from the condition of the damaged road surface, both in terms of structural and functional conditions.
The existing condition of Muara Teweh-Puruk Cahu road section with ± 100 km long is mostly uphill and bends. The road section is currently experiencing significant damage, both minor and severe damage, on several roads and almost along the road section. This road damage disrupts the traffic, both heading to PurukCahu and vice versa.
This study aims to determine the type of road damage on MuaraTeweh-PurukCahu road section and to analyze the correlation between damage and vehicle speed on roads with flat and hilly terrain. The benefit of this research is to provide findings obtained from the results of evaluation and analysis that will later be used as input to the relevant technical agencies, so that the damage repair can be conducted optimally and efficiently.

Criteria of Road Flexibility Pavement
In order to provide a sense of security and comfort to road users, road pavement construction must meet the traffic and strength/structural requirements.

Analysis of Road Performance
Pavement performance of a road section must be able to provide safe and comfortable services according to the plan for the road durability. The pavement evaluation will record the characteristics that can describe the performance of the pavement through several indexes. Based on the characteristics that have been surveyed, pavement evaluation can be classified into functional evaluation and structural evaluation (Christopher Bennett, 2007).
Based on MKJI 1997, the parameters of road performance include the volume and the speed of vehicle.

Road Damage
Road damage is one of the parameters to determine the performance of a road section. Damage on the pavement can be seen from the condition of functional and structural damage. Functional damage occurs when the pavement cannot function as what has planned. Meanwhile, structural damage can be seen by damage of one or more parts of the road pavement structure. Functional failure occurs when the pavement no longer functions as what has been planned and causes discomfort for road users. Meanwhile, structural failure is characterized by damage of one or more parts of the road pavement structure due to unstable subsoil, traffic loads, surface fatigue, and environmental conditions (Yoder, 1975).
One of the ways to find out the type of road damage is by looking at the classification of road damage (BinaMarga) according to the Road Maintenance Manual Number 03/MN/B/1983 issued by the Directorate General of BinaMarga.

Statistics Testing 2.5.1 Regression
Analysis with regression method has two variables, namely dependent variable (Y) and independent variable (X) which have the basic form Y = f (X). Modeling may be influenced by more than one independent variable and the possibility of a large number of independent variables that can affect the program together or separately.

Correlation
Correlation test is used to determine whether a variable has a correlation or affects a problem or other variables.

T -Test
This test is intended to test the independent variable (regression coefficient) whether it has an influence on the dependent variable.

F -Test
Testing on the distribution of F or F-test is intended to determine whether the variables that predict the formation of regression meet the requirements seen from a significant value at a certain level of confidence. This significant value is gained by comparing the calculated F value with F value of the table with a certain level of confidence. Said to be significant if the calculated F value is greater than the table F value.

III. RESEARCH METHOD 3.1
Steps of Data Collection There are 2 (two) types of data that will be used for analyzing this study, namely: a.
Secondary Data Secondary data is data that already exists. In this study, the sources of data can be obtained from relevant agencies. The types of data needed to support this study are administrative boundary maps, road network maps, and road status. b.
Primary Data Primary data is a source of research data obtained by conducting direct observations in the field (surveys) including traffic volume data, road damage data, vehicle speed data, and road geometric data. c.
Instruments The instruments used in this study are in the form of a daily survey form, speed gun, stopwatch, stationery, and other supporting tools related to this study. d.

Procedure of Data Collection
The data needed in this study can be grouped into three data groups, namely traffic volume data, average speed data, and road damage data.

3.2
Data Analysis Technique Based on the data collected, the techniques of data analysis applied in this study are: a.
Analysis of the type of road damage using the BinaMarga method, namely by calculating the and on roads with hilly terrain occurs at Sta. 62 + 300, that is equal to 40.32%.

Vehicle Speed
Based on the local speed survey (spot speed) on roads with flat terrain, it is known that the minimum speed of light vehicle (LV) occurs at Sta. 90 + 400 of 12.59 km/hour with the types of damage that occur are cracks, grooves, potholes, and depression. Based on the calculation of the local speed (spot speed) of heavy vehicles on roads with flat terrain is the minimum speed of heavy vehicles (HV) occurs at Sta. 90 + 400 of 13.14 km/hour with the types of damage that occur are cracks, grooves, potholes, and depression.
Based on the local speed survey (spot speed) when the light vehicle goes uphill, it is seen that the minimum speed of light vehicles (LV) occurs at Sta. 85 + 500 of 18.11 km/hour with the types of damage are cracks and depression. Meanwhile, when light vehicle goes downhill, the minimum speed of light vehicles (LV) occurs at Sta. 74 + 600 at 11.73 km/hour with the types of damage are cracks and depression. Based on the local speed survey (spot speed), the minimum speed at which a heavy vehicle (HV) goes uphill occurs at Sta. 40 + 200 of 14.52 km/hour with the types of damage are cracks, grooves, potholes, and depression. Meanwhile, when heavy vehicle goes downhill, the minimum speed heavy vehicles (HV) occurs at Sta. 85 + 500 of 15.46 km/hour with the types of damage are cracks and depression.
Based on the explanation above, the road performance, in this case is the speed of the vehicle, is affected by road damage, where the dominant ones occur are cracks, grooves, potholes, and depression. Therefore, a statistical test is needed to determine the correlation between the dominant road damage and vehicle speed. Regression normality test shows the histogram graph that the curve line is normal (mean≈0) and the graph of normal probability plots shows that the points tend to approach diagonal lines, so it can be said as normally distributed. Statistical normality regression test of One Sample Kolmogorov Smirnov D count < D table or 0.072 < 0.165, Z count < Z table or 0.594 < 1.96 (95%) and the significance value (Asymp. Sig. 2-tailed) of 0.873> 0.05, then the residual value has been normal. In the multicollinearity test, the Correlations table shows the entire correlation coefficient (R) <Rvalue of the model. Thus, it can be concluded that there is no multicollinearity problem in the regression model.

Analyzing Data using Statistic Testing
Heteroscedasticity test can be done graphically using Spearman's rho correlation coefficient test. Graphically, based on the results of the linear regression analysis plot, it can be seen that the points do not have pattern and spread above and below the axis y (number 0). Thus, it can be concluded that there is no heteroscedasticity problem in the regression model. Based on Spearman's Correlations rho table on the SPSS 20 results, it can be seen that the Sig.  Regression normality test shows the histogram graph that the curve line is normal (mean≈0) and the graph of normal probability plots shows that the points tend to approach diagonal lines, so it can be said as normally distributed. Statistical normality regression test One Sample Kolmogorov Smirnov shows D count < D table or 0.064 < 0.167, Z count < Z table or 0.523 < 1.96 (95%) and the significance value (Asymp. Sig. 2-tailed) is 0.947> 0 .05, then the residual value has been normal.
In the multicollinearity test, the Correlations table shows the entire correlation coefficient (R) <Rvalue of the model, meaning that there is no multicollinearity problem in the regression model.
Heteroscedasticity test can be done graphically using Spearman's rho correlation coefficient test. Graphically, based on the results of the linear regression analysis plot, it can be seen that the points do not have pattern and spread above and below the axis y (number 0). Thus, it can be concluded that there is no heteroscedasticity problem in the regression model. Based on Spearman's Correlations rho table on the SPSS 20 results, it can be seen that the Sig. (2-tailed) Unstandardized Residual > 0.05, meaning that there is no heteroscedasticity problem in the regression model.