Risk Identification and Analysis of Communication Project Based on Fault Tree: The Case of the Telecom IVR Project

Transcription

1 Risk Identification and Analysis of Communication Project Based on Fault Tree: The Case of the Telecom IVR Project BAI Xu School of information Engineering, Wuhan University of Technology, Wuhan, 7, P.R.China Tel: Abstract Aiming at the characteristics of communication s such as the circumstance being complex and affecting factors being numerous, a risk identification method has been put forward based on the fault tree analysis and has combined work breakdown structure(wbs) with structure of risk analysis. With the introduction of WBS-FTS, the coupling judgment has been carried out on the work breakdown structure and risk resource breakdown structure of the telecom IVR and the corresponding risk factors and risk events have been stated. Then the logic connection among each factor has been set up according to fault tree and the upper layer and the lower layer have been connected with logic gate, and thus come to the main fault tree of IVR communication. On this basis, the sensitivity analysis on risk factors has been conducted and corresponding preventive measures have been put forward. Keywords communication ; fault tree analysis; risk identification; sensitivity analysis INTRODUCTION With arrival of the era marked with the integration of data and information, telecom value-added services are also entering the real golden age, and "to enter the field of value-added services" has become a unique landscape. The IVR (Interactive Voice Response) is a new hotspot and benefit growth point within the field of telecom value-added services. IVR communications is a complicated system as well as a reciprocal chiasma interrelated entirety, involving the combination of new technologies and original business, but also including the impact of market conditions and other aspects. Therefore, IVR is a risk and it is necessary to identify and analyze the risk events as well as taking appropriate preventive measures to avoid accidents. At present, domestic and international research on the risk of IVR is still in its infancy, and the risk identification methods mainly include case analysis, factor analysis, SWOT analysis, brainstorming, Delphi method, etc. However, these methods have their own advantages and effects for the identification of risk, but not fixed for the identification of risk. The absolutely effective method should be chosen flexibly according to characteristics and considered comprehensively, thus risk analysis can be conducted quickly and efficiently. Therefore, in this thesis, the WBS method will be used to give system decomposition of the IVR communication engineering work structure and risk source, and couple to the relevant risk factors and events. Then establish the IVR communication engineering fault tree with the logic gate according to the logic connection, and on this basis determine the key risk factors and put forward preventive measures through sensitivity analysis. I. FAULT TREE ANALYSIS Fault Tree Analysis (FTA) is a technology suitable for the reliability and safety analysis of large and complex system. With the application of mathematical logic methods, FTA take the most undesired event of the SciRes.

2 system as the top event of the fault tree, and then look for the trigger event, the direct and indirect causes leading to the event gradually level by level as well as analyzing the logical relationships between the causes of each event. Consequently the various compound modes and the occurrence probability of the causes leading to the system failure are determined, and corresponding measures are taken to improve the system reliability. The characteristics of fault tree analysis are intuitive, intelligible, clear and logical, and qualitative analysis can be made as well as quantitative analysis. Not only the relationships of failures inside the system are reflected, but also reflect the influence of the external factors of the system (environmental factors, human errors, etc.) The process of fault tree analysis is an in-depth understanding process of the system, which requires the researchers to grasp the internal relations of the system and to ascertain the impact way and degree of the underlying factors having on the failures, So that the problems can be analyzed and solved in the analysis process to improve the system reliability. The analysis steps are listed below: Table - The basic steps of FTA the basic Specific contents analysis steps.familiar with Learn more about the system state and the system various parameters, and draw the flow chart or lay-out diagram. Collect incident cases, conduct accident statistics, and envisage the possible accidents. Investigate in the given system. the accident Give comprehensive analysis on the accident investigated to find out the consequences.determine which are more likely to occur and may lead the top event to serious consequence as the top event According to lessons learned and accident cases, solve the accident probability.determine (frequency) as a target value to control the the target value incident by statistical analysis. Investigate all the accident-related cause 5. Investigate events and various factors. the cause Find out the direct cause accident step by accident step from the top event, until get the desired depth of analysis, and draw the fault tree 6. Draw the according to the logic. fault tree Simplify fault tree in the structure,and determine the structure importance degree of basic events 7.Analyze Determine the occurrence probability of all incidents, mark on the fault tree, and find the 8.Accident occurrence probability of the top event probability (accident). Compare and discuss the maintenance 9.Compare system and non-maintenance system. Compare the former and calculate top event probability of the latter..analyze In the analysis, keep flexible according to the particular issue. If the fault tree is large, we can make use of computer. As a logical causality diagram, the fault tree diagram displays the system status (top event) in accordance with the component unit status (basic events). The systematic fault tree logic model formed by Boolean logic gates (such as "and", "or") describes how the combination of the failure and human error lead to the top event. II. THE RISK IDENTIFICATION AND ANALYSIS OF IVR COMMUNICATION PROJECT A. The risk factor identification of IVR communication The risk identification of telecom has made use of the existing achievements such as the telecom IVR flow chart (Figure), the requirements specification and the development plan. Based on the Work Breakdown Structure (WBS) table and the schedule plan of the development plan, the team members can consider the risk factors in accordance with the different stages and each unit in the work breakdown structure diagram, thus forming the risk identification list. According to the risk identification feature of SciRes. 68

3 all-the-people-involvement and in reference to the Delphi method, the risk identification is conducted based on the preliminary risk classification standard and risk lists by the experienced employees from the telecom subsidiary design institute, the center director of the IVR, the network infrastructure department, the information transfer department, the system software control department, the international liaison department and other related sectors, and in this basis a formal risk list document is formed (table -). Through the staff's view of risk identification, calculate the average, and "average" means the average number of days delayed. In addition, the optimistic value, pessimistic value and the most probable value of the risk occurring probability from the staff are converted into a risk occurrence "probability" with the three point method. B. construct the WBS-FTA diagram of telecom IVR se :Overall solution design se:network se : Equipment Project approval risk Market forecasting Design risk Network adjustment Equipment introduction Material supply risk Base station location Base station negotiation From the development process of telecom IVR, the main factors that have led to the failure of the exist in almost every phase and the number of factors is pretty large, so the work breakdown structure (WBS) of the IVR and the FTA analysis combined to evaluate the risk of the telecom IVR. Firstly, adjust Table - by classifying the delayed days of several stages caused by a risk to the first place where the risk appears, and the adjusted table is shown as Table -. Secondly, the fault tree diagram of the telecom IVR constructed in accordance with the basic principle of fault tree and the work breakdown and development risk factors of the telecom IVR development is shown as Figure. Combine table (data derives from survey) and the characteristics of telecom IVR with the fault tree, and give collation and calculation to the data, then it comes to the probability and weight of the risk source of telecom IVR (the basic events on fault tree), shown in table -. Weight is calculated as: R is k. se:the and debugging of Project phase flow FIGURE THE TELECOM IVR PROJECT PROCEDURES AND RISK IDENTIFICATION TABLE - THE TELECOM IVR PROJECT RISK LIST AND RISK Risk factor Overall solution design Network The total process of the IDENTIFICATION se Risk factor s impact on schedule(unit:day) se se Personnel risk Management risk Risk factors of the se ave rag e pro babi lity % SciRes.

4 Excessive design of subscriptions Low level of the clients technicians The frequent extra demands made by clients The strategic objective of the clients being unclear. Network Central apparatus room dissatisfies the design requirements The poor quality of communications circuits Network adjustment affects the users use. The adjustment of network coverage area Equipment Inappropriate Device picking for its unfamiliarity to the manufacture The negotiation risk of base-station brought by market competition the delay of the materials supply The and debugging of software system Cross-departme nt coordination The rework modification of the Slow market reaction of the TABLE - THE TELECOM IVR PROJECT RISK ADJUSTMENT TABLE ris k nu mb er Risk factor Overall solution design M. Inaccurate network capacity survey X. Excessive design of subscriptions X. Low level of the clients technicians X. The frequent extra demands made by clients X.5 The strategic objective of the clients being unclear. X.6 Cross-departm ent coordination X5.7 Slow market reaction of the X6 Network M. Central Apparatus Room dissatisfies the design requirements. X7. The poor quality of communication s circuits X8. Network adjustment affects the users use. X9. The adjustment of network coverage area X Risk factor s impact on schedule(unit:day) P h a se Ph ase Ph ase Ph ase av er ag e p r o b a bi li ty % SciRes. 6

5 Equipment M X..66 X Inappropriate Device picking for its unfamiliarity to the manufacture X. The negotiation risk of base-station brought by market competition X. the delay of the materials supply X The and debugging of software system M. The rework modification of the X X..999 X..88 X X6..88 X7..66 X8..99 X X..555 X..777 X..555 X X..99 C. The risk analysis on the telecom IVR based on Fault tree According to the analysis on the telecom IVR fault tree, the minimal cut set of the telecom IVR can be obtained, and then through the analysis of probability and weight M, M, M, M, T are calculated(shown in Table -. TABLE - THE QUANTITATIVE ANALYSIS ON THE TELECOM IVR PROJECT Basic event Probability q Weight w M T X..666 FIGURE THE TELECOM IVR PROJECT FAULT TREE X..66 X..777 :the delayed days caused by risk factor i of the X..999 X..88 X M=. : the delayed days caused by all the risk factors X6..88 X7..66 T=. -5 of the X8..99 M=.8 TABLE - THE PROBABILITY AND WEIGHT OF THE BASIC EVENT IN THE TELECOM IVR PROJECT PROCEDURES AND RISK X X..555 X..777 Basic event Probability q Weight w X..555 M=.7 X..666 X SciRes.

6 X..99 M=. () The analysis of probability importance degree and criticality importance degree To get to know how much impact the change of basic event s(xi) probability has brought on the top event(t) further, it is expected to analyze the probability importance degree of the basic events, the formula is as: Calculating Ig(): T I g ( i) q Ig()= w*m*m*m=.8 Calculating Ig() Ig() Ig() Ig() Ig (5) Ig(6) Ig(7) Ig(8) Ig(9) Ig() Ig () Ig() Ig() Ig()in sequence, shown as table 5-7 The probability importance degree tells us materiality level of the basic events on the top event, but it failed to reflect on how to reduce the probability of the basic events with a large probability. Hence it is expected to calculate the criticality importance coefficient, which measures the important standard of each basic event from is the point of view of probability and sensitivity. The formula is: LnT CI g ( i) Lnq TABLE -5 THE PROBABILITY AND CRITICALITY IMPORTANCE DEGREE Basic event OF THE TELECOM IVR PROJECT BASIC EVENTS probability importance degree i criticality importance degree X X X X.8-5. X X X X X X X X X X X D. The risk analysis result of telecom IVR ()According to the fault tree of telecom IVR, there are a total of 8 minimal cut sets in IVR, which indicates that there are 8 ways leading to the occurrence of the top event in the telecom IVR. The more the minimal cut sets are, the more dangerous the system is, that is to say that the minimum cut set reflects the risk level of the system. ()From the minimum path set of the telecom IVR fault tree, IVR has a total of minimal path sets, which are the sufficient and necessary conditions to ensure the top event not to occur. There are 8 minimal cut sets and minimal path sets in the IVR fault tree, so it is more convenient to analyze the minimal path set. For example, P = (X) means X does not occur, and then the top event does not occur. And during the analysis that P does not occur, the other can be ignored. That is, if the rework succeeds after the ends, the telecom IVR development can be considered successful even if the se, se and se are not going well. () The probability importance degree of telecom IVR is the most effective approach to prevent the failure of the system development. According to the probability importance degree, the sorting of the telecom IVR basic event s probability is as follows: X> X = X8> X> X = X> X> X> X = X5> X7 = X9> X> X = X6 This indicates that the event that rework modification can not be completed affects the telecom IVR risks most, followed successively by the following events: inaccurate network capacity survey, the poor quality of communication circuits, inappropriate Device picking for its unfamiliarity to the manufacture, the negotiation risk of base-station brought by market competition, SciRes. 6

7 the delay of the materials supply, the frequent extra demands made by clients, the adjustment of network coverage area, low level of the clients technicians, cross-department coordination, central apparatus room dissatisfies the design requirements, network adjustment affects the users use, excessive design of subscriptions, the strategic objective of the clients being unclear, slow market reaction of the. () The criticality importance degree of the telecom IVR reflects each basic event s sensitivity to the risk, and the sorting is as follows: X> X> X8> X> X> X> X> X9> X> X> X7> X5> X> X6> X According to the sorting, the rework modification is the most important for the IVR. Meanwhile, through the comparation of the criticality importance degree and the probability importance degree, it is concluded that the risk of the IVR can be greatly reduced by improving the following problems substantially: if the IVR s, Inappropriate Device picking for its unfamiliarity to the manufacturer, poor quality of communication circuits, the delay of the materials supply, the frequent extra demands made by clients, the adjustment of network coverage area, network adjustment affects the users use. III. CONCLUSIONS The implementation process of the IVR is a complex, one-off, open and complex system involving many relations and variables, and the risk factors affecting the system are numerous and anfractuous. At the same time, the severity of consequence caused by different risk factors is quite different, and there is great uncertainty whether the will achieve the desired result. In this thesis, correlation analysis has been conducted on the telecom IVR with the fault tree method, and the results indicate that: the core of the IVR management is risk management, the rework modification of the, selection when manufacturers do not thoroughly understand; the poor quality of communication circuits, the delay of the materials supply, the frequent extra demands made by clients, the adjustment of network coverage area, network adjustment affects the users use, are the key risk factors in the telecom IVR that call for great attention. Therefore system thinking of the risk management in these areas needs to be strengthened and corresponding preventive measures should be taken to avoid the occurrence of the risks. REFERENCES [] Lau, H, etc. Improved Moment-Estimation Formulas Using More Than Three Subjective Fractals [J]. Management Science, 998. [] Myung J. Maximum Entropy Aggregation of Expert Predictions [J].Management Science, 996. [] Claudio Moraga, Chong fu Huang. A Fuzzy Risk Model and its Matrix Algorithm [J]. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems, (). [] Foreword. Financial Decision Models in A Dynamical Setting [J].Journal of Economic Dynamics & Control,. [5] Li Lin, Li Shucheng, Wang Daoping. Based on Risk-Analysis Estimating Method of Project s Duration [J]. Systems Engineering, (5). [6] Shen Guozhu. Fuzziness Analysis Method for Risk [J]. Systems Engineering and Electronics, (). [7] Shi Xiaojun, Ren Zhi an. A Study on the Risk Analysis of Project Investment: A Method of Influence Causal Reference Based on a Supporting Analytical Algorithm [J]. Systems Engineering-Theory & Practice, (). [8] Jia xiaoxia, Yang naiding. The Supporting Function of Project Management for the Profit of Risk Investment [J]. Scientific Management Research, (). [9] Xie Shengqiang, Zou Huiwen, Chen Demian. Study of the risk distribution and the optimal risk control management in venturecapital investing s [J]. Journal of Anhui University, (). [] Zhang Yali, Yang Naiding. A Project Risk Management Process and Integration under the view of Knowledge Management [J]. Industrial Engineering and Management, 6(6) SciRes.