Primary data were obtained from the results of a field survey which included geometric data, traffic flow for one day on Saturday at 15:00-24:00 WIB, the flow of vehicle entry and exit zones on Jalan Cipto Mangunkusumo, speed survey data, signalized intersection capacity data and unsignalized intersections (U-Turn), and data on the varying capacity of Cipto Mangunkusumo road segments. Secondary data were obtained from the Communication, Informatics and Statistics Office and also the Cirebon City Transportation Office.

Data Analysis Using a form in Microsoft Excel by following the analysis of the 2023 Indonesian Road Capacity Guidelines. The calculation results are obtained in the form of capacity (C), vehicle flow (q), degree of saturation (Dj), queue probability (Pa), and delay (T) to determine traffic services (Level of Service). While the analysis using PTV VISSIM produces speed data, average queue length, maximum queue length and delay (T) which can classify the level of traffic service "LOS" (Level of Service). Based on the Characteristics of Service Levels according to Permendag No. 96/2015 can be seen in Table 1.

Table 1. Service Level Based on Delay (T)

The vehicle growth rate factor (projection) is used to determine the situation in the coming year in order to have alternative solutions that can be done to avoid congestion. Based on data from the Central Statistics Agency (BPS) of West Java, vehicle growth in Cirebon City has increased every year, proving that the increase in vehicles will affect the volume of vehicles on U-turn maneuvers on the road. The steps to calculate the growth rate factor for each vehicle are by using the Compound Annual Growth Rate (CAGR) method with the following equation.

𝟏

𝐕𝐚𝐥𝐮𝐞

( ) − 𝟏) × 𝟏𝟎𝟎

𝐕𝐚𝐥𝐮𝐞𝟐𝟎𝟐𝟎

The projected traffic volume can be modeled in the VISSIM application which aims to obtain the Delay value in producing the traffic level of service (LOS), so that it can compare the performance of the U-Turn with the reference standard Guideline No: 06/BM/2005 concerning U-Turn Planning. The input data are geometric, traffic volume based on incoming zoning flow, speed, vehicle

route, intersection, conflict area, and driver behavior on Jl. Cipto Mangunkusumo.

Road geometric data was acquired by a direct survey using a walking distance meter along Jalan Citpo Mangunkusumo.

Table 3. Road Segment

Table 4. Location U-Turn

Table 3 shows Segment C as the longest at 460 m. From the north, Segments A–D have the widest roads at 9.5 m, while from the south, Segment E is widest at 16.5 m. The total length of Jalan Cipto Mangunkusumo is 2,730 m. Table 4 shows the largest median opening at U-Turn 4, with a width of 27 m, an opening length of 17.5 m, and a closing length of 9.5 m. U-Turns 1 and 2 follow with 21.5 m widths; U-Turn 1 has an opening of 7 m and closing of 14.5 m, while U-Turn 2 has an opening of 12.5 m and closing of 9 m. U-Turn 3 ranks third with an 18 m opening and 14.5 m closing.

In the 2023 Indonesian Capacity Guidelines, U-turns are categorized as conflict movements and are not regulated by the Traffic Signaling Device (APILL), so the approach used uses an unsignalized intersection analysis because this maneuver occurs without traffic light regulation, involves waiting vehicles, and turns against traffic flow. Adjustments to the unsignalized intersection capacity calculation formula add a U-turn factor (FPBa) based on actual conditions in the field, where the U-turn maneuver significantly affects the capacity and performance of the intersection, with the following adjustments (Jenderal et al., n.d.).

C = C0 x FLP x FM x FUK x FHS x FBKi x FBKa x FPBa x FRmi Information:

Table 8. U - Turn Performance 2025

Based on table 8. It shows that U-Turn 2 does not issue a delay time due to the limitations of the 2023 PKJI formula in issuing a delay time when DJ

reaches 1.41, this exceeds the validity limit of the delay calculation therefore, cases like this should be carried out using a microsimulation method such as VISSIM. So the highest delay value is at U-Turn 5 at 21.17 seconds with LOS C, the second value is at U-Turn 4 at 20.59 seconds with LOS C, the third value is at U-Turn 1 at 18.62 seconds with LOS C, and the last value is at U-Turn 6 at 14.61 seconds with LOS B.

The projection analysis uses vehicle growth analysis, assuming that traffic growth is directly proportional to vehicle growth. Vehicle volume growth is calculated using the compound annual growth rate (CAGR) method, which has the same principle as the Compound Annual Growth Rate (CAGR). This method allows for more accurate and stable traffic growth projections over the next 10 years, thus supporting effective road capacity planning and infrastructure management. (Source: Directorate General of Highways, 2012; Road Performance Study, 2022).

Table 9. Cirebon City Vehicle Growth Data

Based on table 9. Shows the values used for the projection are the CAGR of each type of vehicle with a value of passenger cars of 1.23%, buses and trucks are classified as medium vehicles with a value of -1.10% and motorbikes of - 0.08%.

Table 10. U - Turn Performance Projection 10 Year Ahead

Based on table 10. Shows the highest delay value at U-Turn 5 of 29.35 seconds with LOS D, U-Turn 4 of 28.22 seconds with LOS D, U-Turn 1 of 23.59 seconds with LOS C, and U-Turn 6 of 16.96 seconds with LOS C. With these data, it results in a decrease in road performance at U-Turn 4 from LOS C to LOS D, U- Turn 5 from LOS C to LOS D, and U-Turn 6 from LOS B to LOS C.

Based on the analysis using PKJI 2023 in the projection year 2035 regarding the performance of U-Turn 1 to U-Turn 6, an alternative solution is needed to improve the level of traffic service. The modeling approach uses PTV VISSIM software with projected performance data in 2035 and the input is attached in the discussion of the Methodology in microscopic simulation modeling.

Table 11. Parameter Driving Behavior

VISSIM calibration was conducted to modify the characteristics of vehicle drivers according to the site conditions at the study site. Afterwards, the GEH statistical test (Geoffery E. Havers) was conducted; GEH is the standard way to compare two sets of traffic volumes from field data to modeled data. Model validation was conducted by using the GEH test to compare simulated traffic volumes with observed traffic volumes where a GEH value <5 indicates a valid model. (Pratiwi & Rusmandani, 2025).

Figure 10. VISSIM Simulation Results Projection for the Next 10 Years

Based on Figure 10. Showing Traffic Phenomena in the VISSIM simulation in 2035, the image shows the U-Turn 1 location, which is experiencing conflict due to a U-turn and turning onto Jl. Tentara Pelajar.

Table 13. VISSIM Simulation U - Turn Performance Projection 10 Years Ahead

(km/Hour) Intersection Delay (T)

20,07 32,26 - - 29,84 28,45 15,22

Based on table 13. Shows the level of traffic service at U-Turn 1 with a delay value of 20.07 seconds with LOS C, U-Turn 2 of 32.26 seconds with LOS D, U-Turn 4 of 29.84 seconds with LOS D, U-Turn 5 of 28.45 seconds with LOS D, and U-Turn 6 of 15.22 seconds with LOS C.

Alternative 1

Urban roadway median opening requirements have separate values for each class of vehicle passing through it; 4.5 meters for small vehicles, 5.5 meters for medium vehicles, and 12 meters for heavy vehicles. That indicates that the opening sizes are at the respective standards,; however, U-Turn/median opening 6 was only stated to accommodate medium vehicles and has a size of 7 meters. (Pedoman U-Turn.Pdf, n.d.).

Based on Figure 11, Simulation Alternative 1, the researchers made several U-Turn performance coordinations.

Table 14. Alternative Simulation 1

Based on table 14, it shows that the level of traffic service at U-Turn 1 has a delay value of 15.97 seconds with LOS C, U-Turn 2 has a delay value of 10.68 seconds with LOS B, U-Turn 4 has a delay value of 9.43 seconds with LOS B, U- Turn 5 has a delay value of 25.13 seconds with LOS D, and U-Turn 6 has a delay value of 15.14 seconds with LOS C. The data shows an increase in performance at U-Turn 2 from LOS D to LOS B and U-Turn 4 from LOS D to LOS B.

Alternative 2

For the second alternative, do the same thing as alternative 1, but the difference is that engineer U-Turn 5 and U-Turn 6 which have less than optimal performance.

In Alternative Simulation 2, as depicted in Figure 12, the researchers evaluated a number of U-Turn performance optimizations.

Table 15. Alternative Simulation 2

Based on table 15. shows the level of traffic service at U-Turn 1, the delay value is 15.15 seconds with LOS C, U-Turn 2, the delay value is 9.10 seconds with LOS B, and U-Turn 4, the delay value is 9.11 seconds with LOS B. The data shows an increase in the performance of U-Turn 2 from LOS D to LOS B, U-Turn 4 from LOS D to LOS B and closes the service performance of U-Turn 5 and U-Turn 6.

The third alternative essentially involves doing what Alternative 1 did, except for Loopbacks 5 and 6, which did not work optimally.

Based on Figure 13, Simulation Alternative 3, the researchers took several steps to coordinate U-Turn performance, namely.

1. Closing all U-Turn 6s in front of gas stations to optimize traffic service performance, taking into account the existing condition of U-Turn 6,

which is located near a signalized intersection prone to conflict, and the relatively narrow road section at that location for maneuvering into the opposite lane, potentially causing vehicle movement.

Table 16. Alternative Simulation 3

Based on table 16 shows the level of traffic service at U-Turn 1, the delay value is 15.61 seconds with LOS C, U-Turn 2, the delay value is 10.45 seconds with LOS B, U-Turn 4, the delay value is 9.88 seconds with LOS B and U-Turn 5, the delay value is 27.73 seconds with LOS D. The data shows an increase in the performance of U-Turn 2 from LOS D to LOS B, U-Turn 4 from LOS D to LOS B and the closing performance of U-Turn 6.