Production is described as the gradual conversion of one form of material into another form by a chemical or mechanical process to create or improve the utility of the product for the user. It is the activity by which resources, circulating within a defined system, are combined and processed in a controlled manner to provide added value in conformity with policies communicated by management [1].

Productivity is commonly defined as the ratio between the outputs to inputs [2], it measures the relationship between outputs such as goods and services produced, and inputs that include labor, capital and other resources.

Ndungu et al. [2] emphasizes that productivity is an essential factor of economic prosperity: higher productivity leads to higher production and the standard of living in the country is directly proportional to the country’s production levels. With more productivity, a country is able to export and earn foreign currency that amortizes its exchange rate.

Lanna et al. [3] describes productivity as an overall measure of the ability to produce a good or service. Productivity can also be defined as an index measuring output (goods and services) relative to the input (labor, materials, energy, etc., used to produce the output).

Simulation is widely used for process improvement analysis and as a tool to solve many problems in manufacturing leading to improvements efficiency, cost reduction and increased profitability. A simulation is a model that imitates reality. Basic steps in a simulation study include; formulating the problem, building the simulation model, running the model and analyzing the output [2]. Some improvements methods are based on a mathematical description of the process function to be optimized. However, it is difficult for automated manufacturing systems that manage in real time to describe complex control functions using mathematical functions.

B. O. Svensson et al. [4] Described simulation-based optimization as a method to be used when we have a complex system. For this purpose, it is necessary to use simulation to evaluate the performance for each set of parameter values. States that an essential feature of simulation-based optimization, compared to mathematical programming, is that of simulation the objective function is not explicitly known, in fact, this function is defined by the simulation model.

Furthermore, Spieckermann et al. [5], in their study developed a method for improvement for a body shop in the automotive industry. The research was focused on the body shop which requires to be set up every year, therefore the main issue was to find an efficient method to fulfill the layout and optimize the cycle time. In this study, the improvement was done based on a mathematical formulation and an automated optimization approach. The model was based on metaheuristics and genetic algorithm to improve the body shop. The result showed that metaheuristics combined with genetic algorithm were able to detect solutions that cannot be done manually.

Bin et al. [6] in their study addressed a small set of industrial issues using arena simulation. The simulation was based on manufacturing system design at two different stages: supply chain and assembly plant. The simulation was used to reduce manufacturing time and to increase productivity in a company offering spares for light and heavy vehicles in North America.

Furthermore, Jayaraman et al. [7] used discrete event simulation to design automotive powertrain manufacturing systems. The study was focused on a small part of the engine final assembly and repair area. The test was carried on a buffer with engines circulating continuously and entering into a test stand, or a buffer becomes available. The result showed an efficient optimization of the number of stands arrangement. The result showed much time saved using the simulation process and can be implemented for an entire system in the supply chain and production system.

A study by Marsudi et al. [8] used Arena software to simulate a production system of a metal stamping industry. This study focused on evaluating the efficiency and utilization of the production line based on processing time and time utilization in each process; for the evaluation, data was collected through observation in three workstations performing ten different processes to produce a single product.

Another study was based on the production line of a meat processing industry using ARENA software to evaluate the production line [9]. This study was conducted in Turkey to improve the production line due to growing demand. The manufacturing system turns raw meat into a ready to eat the product. Therefore, to improve the production line, different scenarios were used to produce more products using the same working time. The model evaluated the rates of machines and their utilization. The result of this research assisted the company to make the right choice based on different scenarios developed.

In different literature it was demonstrated that production improvement remains a main challenge for automotive industry to be competitive in global market. Therefore, different tools were used for simulation and visualization before implementation in an actual system.

This study is based on a case study from a X truck final assembly line. An analysis of the final truck assembly was made through observation of different process during the production. The problem was to identify bottlenecks in the assembly line to improve product performance. An analysis was carried out, and secondary data were used to develop a conceptual model. The simulation model framework in Figure 1 describes different steps taken to conduct this study.

This section represents the conceptual model of a final truck assembly line. The conceptual model was based on an observation of X truck assembly line. The truck assembly line of the manufacturing processes consists of 12 processes. Three types of data were collected during this study. The first is the layout of the final assembly line of a truck as described in Figure 1.

The process plant map and the processing time was developed using arena software to analyze the throughput of the final assembly line in the automotive industry. The simulation was run for 24 hours using arena simulation software to yield the output of the process assembly line. For easier visualization, the entity was represented with trucks images to demonstrate the manufacturing process and see any queue in the system. Queues demonstrate the station is blocked and process with longer queue has higher processing time. Figure 2 shows the simulation model running in the process. The simulation was made using secondary data and follow an exponential distribution as shown in Table 1. The second data is the operating time in each workstation used to simulate the model. This data is a secondary data collected published by M. dewa et al. [10].

The simulation model has 05 main workstations with an initial output production of 16 trucks day. Through the simulation using ARENA software, the production improved from 16 trucks to 39 trucks, which is 144% improvement in the productivity. To validate the simulation model, a comparison was made using data published by [11] who also conducted research in truck assembly line with also 05 main workstations and simulate using ARENA software have showed that a simulation model using Arena has improved the production to 93% higher than the initial model. Also, the cycle time reduced from 20 minutes per product to 8 min per product. Furthermore [12]. their study supported our

research by improving a truck assembly line using ARENA software and implemented in a real process. Their results showed that the optimal scenario led to a 110% improvement in productivity in the simulation and 60% increase in actual daily production average. Therefore, based on these comparison, we can conclude that the model is valid and imitate a real process.

Figure 3 shows the truck produced in a final assembly line using the initial model. It is observed that 16 trucks were produced during this simulation period.

Figure 4 shows the queue in the system in terms of waiting time and number waiting. The longest waiting time observed is cab preparation and drop, bumper, engine subassembly fitment of soundproof with respective numbers 14.4, 13.97, 13.00, 13.70 minutes, with respective average number waiting of 24.87%, 17.57%, 25.91% and 23.39%. A queue means the process is blocked by the previous one. With this data, we can conclude that cab preparation and drop, bumper, engine subassembly, and soundproof fitment are the constraints in the system.