modeling road traffic accident reporting system by discrete ... - Enim

th

8 International Conference of Modeling and Simulation - MOSIM’10 - May 10-12, 2010 - Hammamet - Tunisia “Evaluation and optimization of innovative production systems of goods and services”

MODELING ROAD TRAFFIC ACCIDENT REPORTING SYSTEM BY DISCRETE EVENT SIMULATION: A CASE STUDY OF IRAN’S ROAD Mohammad Ali Azadeh / Hamid Mohamadlou1 Ahmad Pourahmad/ Saber Mohammadpour2 1 University of Tehran, System engineering department, 2Humane Geography department, Iran [email protected], Hamid [email protected] ABSTRACT In this study business process simulation (BPS) is used to evaluate the effect of redesigning police road traffic accident (RTA) reporting system. It is claimed that by integrating information systems to road traffic accident (RTA) reporting system, duration of this process will be shorten and it will lead to more effective utilization of traffic police personnel. A case is studied in this project is road traffic accident (RTA) reporting system in sample of Iran’s road. In this case the simulation was run ten times in a simulated 30 days for the current road traffic accident system and in presence of new police road traffic accident (RTA) reporting systems. KEYWORD: Discrete simulation, Road Traffic Accident, reporting system INTRODUCTION Improving the work process by applying information technology is one of the most important concerns of organizational management. These redesigns in work process generally take high cost and failures are probable. So for risky attitude of these projects most managers avoid organizational changes. Computer simulation helps manager to study complex systems with providing capabilities such as description of system behavior, scenario analysis, and forecasting before their actual implementation. Simulation can help define deficiencies in the early design process with relatively low cost. A case is considered in this study is road traffic accident (RTA) reporting system in sample of Iran’s road. Two main aspects of performance of the RTA system require improvement. Process execution need to speed up which provide a faster and more efficient service to vehicle drivers and the relatively high staffing cost associated with the process should reduce. This study simulates business process in current state and in presence of information technology systems to estimates potential cost savings. DISCRETE EVENT SIMULATION SYSTEMS In general there are two approaches to computerbased modeling. One approach is referred to as Systems dynamics and the other is termed Discrete Event Simulation. Either can be effective in

analyzing business problems, though as a generalization, Systems Dynamics is usually taught in MBA classes and used by consultants for the strategic analysis of options, while Discrete Event Simulation is typically packaged with business process modeling tools and used to determine how actual processes perform, or to project how they might perform under various future scenarios. In This paper we focus on Discrete Event Simulation. Discrete event simulation models are based on events that occur within the course of a business process. Probabilistic and deterministic discrete event simulation is commonly two types of implementations of model. Most business process work relies on probabilistic discrete event simulation models. These models are used to define the business work steps, specifically the entities that flow through the business, as well as the resources required to perform each work step. A simulation expert must enter information about the flow of events. The timing and occurrence of the events are based on probability distribution functions which reproduce the behavioral dynamics of a real world business process. The developer of the simulation must choose probability functions that reflect the behavior of a given process. The process model and the information about the events are entered into a software program which can then "execute" the simulation. In this paper we use Visual SLAM software to simulate business process. By entering initial data and executing the system, the modeler can determine future states of the process. As the

MOSIM’10 - May 10-12, 2010 - Hammamet - Tunisia

simulation executes, the events are generated, the entities flow through the process, the delays are sampled, and the resources are used-all using probability distributions to produce the real-world randomness of the business process. Consider a simulation of our road accident reporting system process. We have a probabilistic function that determines how many road accidents occur in one sample road. This function is based on statistical data gathered by organization. If we indicate that 100 accidents occur on Monday, our system will automatically assign a portion of them to the related traffic accident police. As the number of accidents increases resources, ranging from police officer and base station staff and HQ department clerk, must be increased. By running different simulations we can determine exactly what specific additional resources will be required to accommodate the varying levels of increase in the number of maternity cases. Stepping back from the process described in the paragraph above, it is obvious that a good simulation will depend on good data about the resources consumed in each activity within a process, the time required for each activity and the employees required accomplishing the activity. Thus, anyone undertaking a simulation will also, in effect, be developing a cost model for the business process that forecasts the costs and times required to execute certain processes over specific periods of time. Thus, for many, simulations are used to determine the cost of new or revised processes. If one is creative about the scenarios one considers and runs simulations with different input assumptions, one can also identify bottlenecks or problems within processes that might not otherwise be identified. This is especially true of complex processes with feedback loops. Most process analysts have no trouble figuring out the logical sequence a process should follow, but few are capable of identifying problems that might occur as the volume of process events increases and feedback begins to occur in real time. Simulation systems are usually the only way to be sure that new processes will work when subjected to a variety of real world situations. Historically, companies that did business process modeling rarely went on to do simulation. It's a lot easier to do a process model than a simulation model. Simulation requires that you know a lot more about the activities that make up a process. You need explicit decision rules to route the flow, cost, time and resource utilization data for each activity. You also need a reasonably good knowledge of simulation. Smart simulation analysts will warn you not to try simulating everything - it is too complex. Instead,

they will tell you to focus on the key things that really make a difference. In addition, good simulation systems offer lots of formulas for characterizing the nature of the flow of data through a process. Learning which formulas to use in which situations is also something you learn from experience. In other words, you need to be prepared to hire a consultant, or invest enough of your own time to really master simulation. BUSINESS PROCESS SIMULATION An important part of the evaluation of designed and redesigned business processes is Business Process Simulation (BPS). Business Process Simulation is attracting attention more than a decade now. Companies are improving their performance by a constant evaluation of the value added in all parts of their processes. Business processes are in a continuous improvement cycle in which design and redesign play an important role. Various possibilities to change a process are present and the best alternative design should replace the current process. Making an intuitive choice may lead to unpleasant surprises and lower process performance instead of yielding the expected gains. The simulation of business processes helps in understanding, analyzing, and designing processes. With the use of simulation the (re)designed processes can be evaluated and compared. Simulation provides quantitative estimates of the impact that a process design is likely to have on process performance and a quantitatively supported choice for the best design can be made. BPS (Tumay, 1996) has been traditionally used in a manufacturing context, but it is increasingly used in service organisations. Levine and Aurand (1994) describe the use of simulation to analyze an automated workflow system of an administration process, Greasley (2000) describes the use of simulation to analyze the custody of prisoner process and Verma et al. (2000) describe its use in redesigning check-processing operations. The use of simulation to assess the implementation of information systems (IS) in particular has been described by Giaglis (1999) and Giaglis et al. (1999) who outline the need to support IS evaluation by developing techniques for generating estimates of the organizational value of IS. Experimental methods such as simulation are suggested as capable of providing such estimates. Warren and Crosslin (1995) suggest simulation as a method of providing the justification necessary to support business decisions to redesign and make what are often massive investments in IT systems. A distinction should be drawn between studies of change to information system process design and change to the


information infrastructure itself, in terms of elements such as the IS network configuration, telecommunications hardware and application software. Changes to these aspects are usually undertaken using network simulation software. Painter et al. (1996) argue that the process and infrastructure analysis can be integrated and present a methodology to achieve this. Also, the importance of the integrated business process, information process, and production process modeling has been highlighted by several recent publications including

Brandon, Betts, and Wamelink (1998), Lin and Shao (2006), Leshchyshyn and Rieb (2004), Khouja and Kumar (2002), Ali, Chuang, and Bernard (2002), Pandolfi (1993), Dewan and Min (1997), Yamazaki and Maeda (1998), Bernard and Perry (2003), Milaev, Fatkin, and Rulyaeva (2002), Swindells (2002), Kalay (2006), Choo and Kim (2002), Grover, Teng, Segars, and Fiedler (1998), and Claussnitzer, Wittmann, and Gruebnau (1999). Some of the most important simulation studies in the areas of BPS, IT, and MIS are categorized in Table 1.

Table1. Some of recent study in BPS, MIS and IT Author

Year

Business process simulation (BPS)

Giaglis, Paul

1997

Giaglis, Paul

1999

Bastos & Ruiz

2001

Gunasekaran Karacapilidis Seshasai, Gupta

2002 2004 2005

Rungtusanatham & Forza Andrew Greasley

2005 2005

Ammar Eich, Fan, Sun

1991 1992

Humphreys, Berkeley Sol & Streng

1992 1992

Warren, Norcio & Stott Proctor

1992

Kobayashi

2002

Su, Yin & Chang Johansson

2002 2004

Hyatt, Contractor & Jones Giaglis, Paul & Okeefe Asseldonk, Jalvingh, Huirnc Mointra& Konda

1996

The simulation of business process in the setting of process design inside the organization of paramedical industry An integrated simulation on organizations design studies, the effect of simulation in companies design processes. Towards an approach to model business processes using workflow modeling techniques in production systems Modeling and analysis of business process reengineering Enhancing collaboration in business process modeling An integrated and collaborative framework for business design: A knowledge engineering approach Coordinating product design, process design, and supply chain design decisions: Part A: Topic motivation, performance implications, and article review process A redesign of a road traffic accident reporting system using business process simulation MIS simulation A system to support the operation analysis of software with high throughput. Methodology for the simulation of database architectures for performance evaluation. A simulation for organizational modeling. An approach in which the inter organizational dynamics are represented in terms of layered actors in network and entities. The simulation models based on probabilistic discrete events of a data flow diagram warehouse and the information which describe the run of system components. Discrete simulation is presented as an efficient tool to find useful solution in office management. An approach to a dynamic system simulation based on human information processing Computer simulations of an integrated distribution information system A system for information management in simulation of manufacturing processes IT simulation An object- oriented simulation environment designed and built to specifically support simulation of organizational networks. The potential of integrating business and network simulation models to facilitate concurrent engineering of business processes and information technology Potential economic benefit in changes management via information technology applications on evaluation Dutch dairy farms by simulation. Simulation model for managing survivable of network information systems. A model for of the relation between expenses and defense mechanism for

1997

1999 1999 2000


Richards

2002

Khouja & Kumar Nystrom & Risch

2002 2004

network system. A simulation model with endogenous technical advance information technology and increasing returns form research. Information technology investments and volume-flexibility in production systems Engineering information integration using object-oriented mediator technology

quickly provide agencies with relevant RTA information. Using PDA and mobile system and other technology such as OCR (optical character recognition) and DIP (document image processing) police force can to facilitate communication .Officer returning to base station will be omitted with use of such technology to send data gathered by police force in RTA scene. It will lead to more cost savings and efficient processes. The current and proposed process designs for actions in road traffic accident will now be discussed. The flow chart of simulation model for case study is shown in Figure 1.

CASE STUDY The system being studied is a sample case in one of the Iran’s road. A sample road is chosen and statistical information for evaluating system is gathered. These statistics is used to run the simulation model in current state. Computerized RTA systems are to both reduce the cost of the process and increase process execution speed. Digital technologies generally used are mostly to be collect data and transmit to central database. Such systems are geographical information system (GIS) that can

RTA

Recording to database Observe by public

Database

Public HQ Department

Inform to station

Getting Information

Base Station

Geographical Locating

Send police Send policeofficer

Officer

Write Abstract Write

Court Write Abstract

Form Distribution Informing

Oversee Witness Evidence

Witness statements

Traffic Administration Witness

Insurance Companies

Figure1. The DFD of road traffic accident business process


THE ROAD TRAFFIC ACCIDENT BUSINESS PROCESS SIMULATION The main stages of the RTA process are commonly executed in Iran’s current system of reporting will now describe. After base stations receive an accident reporting by public, police officers attend in location and fill a number of paper-based forms. These forms are distributed to the traffic administration and data HQ (Headquarters) departments for processing. The traffic administration section oversees the submission of witness evidence, either by post or in person and the collation of an abstract containing officer and witness statements for use by interested parties such as insurance companies and court proceedings. The data HQ section oversees mapping of the geographical location of the accident that is used for road transport initiatives such as traffic calming and speed cameras. Following the notification of a road traffic incident to the police by public, a decision is made to attend e scene of the incident. It may be that for a minor incident the parties involved are instructed to pursue proceedings with their insurance companies and the police have no further involvement. If it is necessary to attend the RTA scene the officer travels to the location of the incident. After an assessment is made of the incident the officer returns to the station to complete and submit the appropriate paperwork. Three forms are used by a Police Officer attending a road traffic accident (RTA). Form type 1 is used for injury accidents contain 3 page. Page 1 and 2 are forwarded by the officer to the witness pro forma process and the page 3 is forwarded to the data HQ section for location mapping. Form type 2 is used for non-injury an incident which is filed unless further action is to be taken as a result of a dispute or claim, when it is then passed to the witness pro forma process. Form type 3, pocket book entries (PBE), are taken when no official record is required but provide data that could be retrieved at a later date and transferred to the appropriate form. Amendment of forms may take place at a later date. Amendments of forms Like further action such as new witness statements will be required if changes take place. Form type 1 page 3 amendments are communicated to data HQ. A location mapping process collates information and passes it to the local council who provide a location grid reference from sketches and location information provided by the officer who attended the accident scene.

The data is collated, sorted and then forms are mapped in batches by entering location codes on new forms. If all the necessary information is not available a memo is sent to the officer for further information and the process is repeated. The witness pro forma process obtains accident witness and driver information and places it on a pro forma sheet. If a witness is identified their details are taken and a pro forma is sent to them. If a fatal accident has occurred then the officer obtains further details in person at a later date. If the pro forma has not been returned after three weeks a reminder letter is sent to the witness. If there is still no response from the witness and the information requested is required for further proceedings, then an officer will obtain the statement in person. The abstract preparation process collates and checks documents associated with the RTA process to ensure all the data needed has been received. A decision is made at this stage if further action is required after reviewing the evidence collected. If further action is required a number of forms are collated. If, at this stage, no prosecution is to take place, a letter informing the driver of this decision is sent. If a prosecution is to take place the officer will write an abstract, summarizing the details of the case. If a court case is scheduled and a “not guilty” plea has been entered then the officer will be required to attend the court proceedings in person. Otherwise this is the end of the involvement of the officer. Process network that describes the logic of the model are shown in the figure 2. Tow kind of data are required to simulate the model. Decision points can be modeled by a conditional rule based method or by a probability distribution. Probability distributions for decision points, such as the proportion of injury and non-injury events are derived from the sample data and take the form of a percentage. Table 2 shows the results. Table2. Some of decision point Probability distributions in simulation Function Need to attend in place Injury More action Fatal accident Prosecution Court case

Probability (TRUE,FALSE) (0.8,0.2) (0.8,0.2) (0.8,0.2) (0.5,0.5) (0.7,0.3) (0.6,0.4)


RTA

Attend?

T

T

Injury?

Attend scene

F

Fill form1

F

T

Terminate

More action?

Fill form2

F Fill PBE

Record DB

Locating

T ‫ﺑﻠﻪ‬ F

‫ ﺧ‬Witness pro forma

Fatal?

T Completed?

‫ﺧ‬

Essential?

T

In division?

Get statement

F

F

Contact division

T More action? ‫ﺑﻠﻪ‬

‫ﺧ‬

F Terminate

Collate forms

Prosecution?

‫ﺧ‬

‫ﺑﻠﻪ‬T

T Write abstract

Court?

F

Write to driver

F

‫ﺑﻠﻪ‬

Guilty?

F T

Terminate

Terminate

Terminate

Figure2. General logical flowchart of the simulation model

Terminate

Attend court


The second area of data required for the simulation model is for additional elements such as process durations, resource availability schedules and the timing of RTA occurrences. In this case probability distributions for process durations are derived from the sample data. In general a triangular distribution has been used for process durations that require minimum, mean and maximum parameter values. Resource availability, in terms of a police officer attending the RTA, is assumed to be infinite as an RTA incident is treated as an “emergency situation” and if the designated officer is unavailable an alternative officer is found. Over a period of six years there had not been an incident when no officer could be found, when required, to attend an RTA scene. The simulation model was developed by Visual SLAM (Pritsker, Oreilly, & LaVal, 1997; Pritsker,

1990; Pritsker,Sigal, & Hammesfahr, 1989) that incorporates a template of shapes called nodes that are placed on the computer screen and connected to represent the logic of the model. Each entity flows in the system by one real attributes (business process information). According to Visual SLAM terminologies the real attribute should be shown by “ATRIB [I],” the integer attribute should be shown by “LTRIB [I],” and the variables should be shown by “XX [I].”XX [2] and XX [3] are temporary (dummy) global variables and are only used for controlling computer programming. All related process duration and rate of accidents functions were stop watched and time-studied or quantitatively identified from experienced police officer. According to Visual SLAM terminologies, exponential,

Table3. Distribution function of some of the RTA process Function Time between road accident Attend in RTA Filling form type1and fax to base Filling form type2and fax to base Filling form type3and fax to base Geographical locating Recording to database Getting witness pro forma Get statement in person Collate forms Letter to driver Write abstract Attend in court triangular, and uniform distribution functions are shown by EXPON (XMN), TRIAG (MIN, XMN.MAX), and UNFRM (ULO, UHI) respectively. It should be noted that XMN, STD, ULO, and UHI are mean, standard deviation, lower limit, and upper limit, respectively. The discrete distribution functions are shown by DPROBN (IRCUM, IRVAL), where IRCUM and IRVAL are arrays of probabilities and arrays of sample values defined by the user. The main elements that make up a model are the “Create”, “Goon” nodes and “Activity” connector. The “Create” node generates the arrival of people (service applications), physical components (manufacturing applications) or information (information system applications) into the system. These are generically called entities in simulation terminology. It is necessary to define the rate of arrival of entities in the system by defining the time between arrivals (inter-arrival rate). This was

Distribution EXPON (24) TRANG(0.3,0.4,0.5) TRANG(0.2,0.4,0.6) TRANG(0.1,0.3,0.5) TRANG(0.3,0.4,0.5) UNFRM (0.75,0.1.25) TRANG(0.1,0.12,0.13) UNFRM (1,2) TRANG(2,3,4) TRANG(1,2,3) TRANG(24,48,72) TRANG(12,24,36) UNFRM (3,5) achieved using sample data on the timing of the occurrence of RTA incidences. The “activity” connector is used to delay an entity for a predetermined time period (to represent the time taken to attend the RTA for example). It is also used to allocate resource time (e.g. traffic officer time) to a process. The “goon” node is used to define decision points within the process. The entity can leave from two or more outputs from the node. These decision options can be implemented as either a percentage chance for each decision route (e.g. 70 per cent entities follow the true route, 30 per cent false) or by a rule-based decision. Before experimental analysis of the model can begin it is necessary to ensure that the simulation provides a valid representation of the system. This process consists of verification and validation of the simulation model. Verification is analogous to the practice of “debugging” a computer program. In this case the first check was to run the


simulation and observe the progress of the entities through the system to check for any logic errors. A verified model is a model that operates as intended by the modeler. However this does not necessarily mean that it is a satisfactory representation of the real system for the purposes of the study. This is the purpose of validation. In this case the results from the current RTA reporting system model could be compared against historical data of the actual system. A decision is made if model behavior is close enough to the real system to meet the objectives of the study. Unlike verification, validation is a matter of judgment that involves a trade-off between the accuracy of measurement required and the amount of modeling effort required to achieve this. Because of the probability distributions used for RTA events, process times and decision points, the output measures of the simulation vary each time the simulation is run. Therefore it is necessary to run the

simulation multiple times and form a confidence interval within which the average of the measure should lie. In this case the simulation was run ten times for a simulated 30 days for the current system. The results show that the mean officer hours required to execute all the tasks associated with the RTA process is 14.755 hours under the current system. The results have shown in table 4. The minimum required time to execute all the tasks associated with the RTA process was 5.9 hours and the maximum was 23.600. While results show that 27 accidents are occurred during 30 days simulation period, the standard deviation of officer involving time in each accident is 2.077. In the second and third run of simulation the mean values are 9.213 and 23.755 hours and number of accident are 31 and 24. 3 unit of police officer is considered in each station and average available police officer in each time was 0.943 for the first run.

Table4. Result of simulation runs for current system Number of Runs

1 2 3

Statistics

Mean Value

Standard Deviation

Number of Observations

Minimum Value

Maximum Value

Current Capacity

Average Available

Police involving time Police resource Police involving time Police resource Police involving time Police resource

14.755 2.057 9.213 1.608 23.755 1.443

2.077 1.076 0.209 2.414 2.077 1.923

27 31 24 -

5.954 3.954 21.954 -

23.600 19.231 26.600 -

3 3 3

0.943 1.578 1.981

THE PROPOSED SYSTEM A data warehouse for storing data with online processing capability, mobile technology such as notebooks, PDA with GIS systems, document image processing (DIP) system, optical character recognition (OCR) are such a technology that can shorten business process of RTA and lead to cost saving. In the proposed computerized RTA reporting system the attending officer completes paper-based forms as before but this information is promptly converted to digital form using a document image processing (DIP) system. This is achieved by a combination of image capture and data recognition through a facsimile link. Data recognition systems, such as optical character recognition (OCR) are used to process information that is entered in a structured format, such as options selected using a ticked box format. Image capture is used in the following ways. Documents are stored as images to enable input bureau staff to validate the OCR scanned data. Images that cannot be interpreted by data recognition software, such as hand drawn sketches of the RTA scene are stored for later retrieval. Images of text, such as officer-written notes, can be entered by input

bureau staff, saving officer time. Once in digital format the documents can be delivered electronically preventing data duplication and enabling faster distribution. Physical documents are held in a central repository for reference if needed. Location details are currently based on a written description of the RTA by an officer which leads to inconsistent results. The current location description by the officer is usually acceptable for city incidents where nearby street intersections and other features can be used to pinpoint a location. However on long stretches of road it is often difficult to pinpoint an exact spot. This is important because of the need to accurately pinpoint areas with high accident rates for road safety measures (e.g. road humps) and speed camera placement. Further inaccuracies can also occur when the officer description is converted by the local council using an Ordnance Survey (OS) map grid reference which is only accurate to 200 yards. In this proposal each officer is issued with a portable digital map on which to indicate the RTA location. This information is transmitted by a mobile link to a geographical information system (GIS) (Wiley and Keyser, 1998) which provides accurate location analysis of both injury and non-injury incidents using


the geocode system (Radcliffe, 2000). The GIS system will combine the accident location analysis with data relating to the location of pelican crossings, traffic lights, street parking and anything else that might contribute to accidents or affect schemes being proposed. Along with data on details on road conditions at the time of the accident this information will help determine a prioritized list of road safety improvement measures. These systems will implement in one of the Iran’s sample road in period of three month and the result will compare with current state. The need for officer time for transcribing and updating notes will be minimized by the use of optical character recognition (OCR) transcription of officer notes by data centre personnel. A centralized data store will also save officer time by quicker storage and retrieval of information for the pro forma and abstract preparation process. In addition the location mapping exercise will be simplified by the use of portable digital maps from which officers can indicate the RTA location. Process time will be speeded by workflow automation software that will prompt for timely response to requests for information in the witness pro forma and abstract preparation processes. The use of a single point of contact for all data submissions and information requests will also reduce process execution time by eliminating search and delivery delays associated

with paper records. The IS system will also have the benefit of improved data accuracy with a single database of all information and location analysis through the use of geocodes. The simulation study focused on savings made on the front-line road traffic officer staff but substantial savings can also be made by the centralization of the traffic administration units. These units are currently located at a divisional level, which is a geographical subdivision of the Police Force area. This is necessary so that paperwork can be processed from officers returning to their local stations. However with the use of digital transmission of RTA information the geographical location of the administrative support can be centralized at a Force level. This can lead to less staff needs due to a centralized automation of processes through workflow and database technologies. The demand on separate divisions would also be aggregated at a corporate level leading to more efficient staff utilization. For implementing the simulation in presence of information system we used expert data and as we discussed before some of bureaucratic process will be discarded in new system and some of them will take shorter time to process by utilization of new technology. As it is anticipated, in the table 5 in 3 runs of simulation in presence of new system, the process time in each one have shorten comparing to the current system.

Table5. Result of simulation runs for new system Number of Runs

1 2 3

Statistics

Mean Value

Standard Deviation

Number of Observations

Minimum Value

Maximum Value

Current Capacity

Average Available

Police involving time Police resource Police involving time Police resource Police involving time Police resource

11.671 1.937 10.241 1.831 13.748 1.576

2.341 0.698 0.963 1.994 2.621 1.703

30 29 25 -

3.857 5.774 6.673 -

20.809 17.231 21.891 -

3 3 3

1.961 1.578 2.034

CONCLUSION This study has utilized simulation techniques to assess business process of accident reporting system at a current state in one of the Iran’s sample road. Also a new system base on using IS and related technology proposed. It was claimed that in proposed system process execution is improved and it was led to savings in officer time and parties. The information system offers a number benefits including faster delivery of road traffic accident information to government agencies through the use of digital mapping technology replacing location analysis by a third party. Faster process execution of

document flows though workflow technology rather than the movement of paper records is also possible. Cost savings in terms of road traffic officer time are gained by eliminating the need to return to the police station to undertake administration duties. Savings are also predicted though a centralized administration facility, enabled through the digitization of data, rather than paper records kept at a divisional level. The quality of the process is improved by the greater accuracy of road traffic accident location analysis through the use of digital mapping and geocodes. Improved accuracy of RTA information will be gained though the use of a centralized database store


replacing paper documentation. The simulation indicated that utilizing information system will lead to both time and cost saving of process.

REFERENCES Andrew ,G.,(2004), A redesign of a road traffic accident reporting system using business process simulation, Business Process Management JournalVol. 10 No. 6, pp. 636-644 Azadeh,A., M. Haghnevis, and Y. Khodadadegan, (2007). Design of the Integrated Information System, Business, and Production Process by Simulation, JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE AND TECHNOLOGY, 59(2):216–234 Azadeh, A., Haghnevis, M., & Kodadadegan, Y. (2005). Assessment of a complex machine mix problem by integrated simulation and AHP modeling. In Proceedings of 5th International Conference on Analysis of Manufacturing Systems—Production Management. Zakynthos Island, Greece. Azadeh, A., Haghnevis, M., & Kodadadegan, Y. (2005). Design and assessment of the integrated information,business process and production system by simulation. In Proceedings of 5th International Conference on analysis of Manufacturing Systems— Production Management. Zakynthos Island, Greece. Azadeh, M.A., & Ebrahimipour, A. (2004). An integrated approach for assessment and ranking of manufacturing systems based on machine performance. International Journal of Industrial Engineering, 11, 4. Fathee, M.M., Redd, R., Gorgas, D. and Modarres, B. (1998), “The effects of complexity on business process reengineering: values and limitations of

modeling and simulation techniques”, in Mederios, D.F., Watson Giaglis, G.M., Mylonopoulos, N. and Doukidis, G.I. (1999), “The ISSUE methodology for quantifying benefits from information systems”, Logistics Information Management, Vol. 2 No. 1, pp. 50-62. Greasley, A. (2000), “A simulation analysis of arrest costs”, J. Opl. Res. Soc., Vol. 51, pp. 162-7. Hyatt, A., Contractor, N., & Jones, P. (1996). Computational organizational network modeling: Strategies and an example. Computational and Mathematical Organization Theory, 2(4), 285–300. Johansson, H., Astrom, P., & Orsborn, K. (2004). A system for information management in simulation of manufacturing processes. Advances in Engineering Software, 35, 10–11, Engineering Computational Technology, 725–733. Jordan, E., & Evans, J.B. (1992). The simulation of IS strategy using SIMIAN in dynamic modeling of information systems. Amsterdam: Elsevie Science Publishers. Kalay, Y.E. (2006). The impact of information technology on design methods, products and practices. Design Studies, 27(3), 357–380. Kelton, W.D., Sadowski, R.P. and Sadowski, D.A. (2001), Simulation with ARENA, McGraw-Hill, Singapore. Levine, L.O. and Aurand, S.S. (1994), “Evaluating automated work-flow systems for administrative processes”, Interfaces, Vol. 24, pp. 141-51. Painter, M.K., Fernandes, R., Padmanaban, N. and Mayer, R.J. (1996), “A methodology for integrating business process and infrastructure models”, in Charnes, J.M., Morrice, D.J.,