Management & Engineering 03 (2011) 1838-5745 Contents lists available at SEI Management & Engineering journal homepage: www.seiofbluemountain.com Applying DEA to Performance the Construction Safety Management Jinxian Zhao , Yuanlu Qiao School of Management, Qingdao Technological University 266520, China KEYWORDS ABSTRACT Construction safety, DEA, Safety management In order to solve the problem of the construction site safety evaluation performance, this thesis come up with one method based on DEA to evaluate it. From this paper, you will find the author try to explain the connotation of Data Envelopment Analysis (DEA) and apply it to the construction management research. Through this way, we can widen the applications area of DEA method. Meanwhile, a sample measuring the On-site safety management is given. © ST. PLUM-BLOSSOM PRESS PTY LTD 1 Introduction Construction is one of pillar industries in china, while safety accidents happened frequently in this area. Therefore, many scholars place a great value on this subject and take the research of construction safety management seriously. At present, China is in the large-scale infrastructure construction. It goes without saying that safety management and construction site safety performance evaluation obtain the same weight. So doing research on the construction site safety performance evaluation has an important practical significance on improving the construction site safety and ensuring the smooth implementation of the construction project. In order to judge construction site safety management effectiveness, we need to evaluate the construction site safety performance. Nowadays, the method of safety checklist is widely used in Chinese construction. [1] However, this method has great limitations .For instance, it can not reflect the overall construction site safety management effectively. And Chinese scholars do a lot research on the construction safety performance, and they aimed at the index of security measure, which is proposed to based on AHP and other methods ,such as analytic hierarchy process (AHP) and grey theory of fuzzy evaluation method [2-3] as combination method to evaluate. These methods are only an assessment of safety performance, and lack of accurate information about weighting factor, and more important is that it can not accurately response building safety management performance of the site. TO evaluate the safety management performance of construction site, in practice, is to judge of the security management in the construction site or is to evaluate whether the output could reach the maximum when the input of security management is kept constant, or is to assess whether the input could stay minimal when the output is kept constant. This provides a similar solution solved by models of Data Envelopment Analysis (DEA). DEA is a perfect method to evaluate the “efficient scale” and “efficient Corresponding author . Email: qiaoyuanlu2006@163.com English edition copyright © ST. PLUM-BLOSSOM PRESS PTY LTD DOI:10.5503/J.ME.2011.03.018 109 technology” of one program which is equipped with multiple input and output events. My paper started with the analyzing of DEA model, and then we studied the established program of security input and loss in the construction with the cases of verification research. In addition, I choose the effectiveness of security management to evaluate the performance of construction security. 2 DEA Model The basic idea of DEA is to form the first decision unit to determine the input and output factors. If n departments involved in one production activity, they can be schools, hospitals, courts, construction projects, and it also can be banks or enterprises. Each DMU has m types of factor inputs and s types of outputs. xij is ith input amount for DMUj, x ij >0, yrj is rth output amount for DMUj, yrj>0,vi is weight of ith input ,ur is weight of rth output ,i=1,2,…m,j=1,2,…n,r=1,2,…s. Recorded input vector as Xj and input vector Yj,v and u is weight vector of m inputs and s outputs, we have X j x1 j , x2 j ,, xmj , T v v1 , v2 ,, vm T , j 1,, n ； Y j y1 j , y 2 j ,, y sj , T j 1,, n u u1 , u2 ,, um T The efficiency evaluation index can be showed hj uTYj vT X j , j 1,2,, n We can choose weight coefficient v and u , let h j 1, j 1,2, , n For the efficiency of DMUjo,we have the target hj0 u T Y0 vT X 0 So , we compose a fractional programming (C2R model) u T Y0 max V PI vT X 0 uTYj 2 I (C R ) 1 vT X j u 0, v 0 Let t 1 , w tv, tu v X0 T (1) , fractional programming (1) problem can be changed into linear programming (2) max T Y0 VCI2 R T X j T Y j 0 ( PCI2 R ) T X0 1 0, 0 (2) DMU j0 is weak efficient if the optimal value of linear programming VCI2 R 1 0 0 and 0, 0 , DMU j 0 ( PCI2 R ) VCI2 R 1 ,if the model ( PCI2 R ) is efficient. Following Charnes and Banker, we assume that the technology sets is estimated by 110 optimal solution n n T ( X , Y ) j X j X , j Y j Y , j 0 j 1 j 1 Correspond to the technology sets, we have the following linear programming problems which called C2R model min VD n j X j X 0 j 1 ( D) n j Y j Y0 j 1 j 0 (3) 3 Construction Safety Management Performance Evaluation Construction safety management investment and benefits not only include construction companies or construction projects for the security of investment, but also the direct and indirect casualty losses. How to apply DEA model to evaluate the effectiveness of building safety, the answer is that we need to do the statistical analysis of their input. According to the survey conducted by Fang Dongping, [5] who divide the safety input into four aspects, individual labor protection, civil construction, safety education and on-site safety. The proportion as the Figure 1: Safety investment proportion 1% 3% 29% 67% Personal labor protection The on-site safety measures Civilization construction Safety education Figure 1 Safety input and proportion As the Figure 1 conclude, on-site safety facility and civil construction takes the largest proportion, and personal protection and safety education work very little. So we select three of them as the DMU units—evaluation index, on-site safety, civil construction, protection of individual labor. At least as safe investment in education, the indicators will not be considered again. Security investment income is received from the accident rate of the resulting casualties and economic losses (including direct economic losses and indirect economic losses). Nevertheless, it is because the DEA evaluation is the evaluation of its effectiveness, and output indicators must be direct benefits. Traditional safe accidents are recorded as losses. Construction accidents and injuries are divided into economic losses, and casualties can be counted as general economic losses. In this paper, we divide the loss of construction into two areas, the direct losses and indirect losses. Through statistical analysis, the more serious casualties, the greater the loss the company did not record, and it can be say the greater the indirect losses that is accidents. Safety management brought about by the accident rate of casualties and economic losses are not suitable for DEA. Here we transforming economic losses into two aspects, one is economic benefits of enterprise bring about by the direct economic loss. The greater the economic losses, the lower the economic benefits of enterprise are. The other is the indirect losses, the greater the indirect loss of business, the greater the impact on the company's social benefits, social benefits and thus lower the corporate. According to DEA model, the output units include social and economic benefits. 111 4 Case Study The effectiveness evaluation of construction safety management should be measured from the logical relationship between safety investment and income from the security (social and economic benefits). Firstly do DEA effective analysis of all executation parts which are deemed as different appraisal unit, and then representative investment items are determined as follows: on-site safety x1, civilized construction x2, x3 personal labor protection as three input x, their income factors: social and economic y2 y1 as two output y. (In dollars / year) Here selecting Qingdao 10 major construction projects to evaluate, by statistical analysis of each item, its index data see table 1 below: DMU Serial Input indicators Output indicators numbers X1 X2 X3 Y1 Y2 1 2000 3355 42 5000 30000 2 1300 2211 57 3562 38000 3 1020 1400 38 2191 20000 4 330 800 39 1643 12000 5 450 750 50 2329 11000 6 960 2110 42 4000 20000 7 2800 1850 49 3562 30000 8 849 1208 56 1233 19000 9 680 1000 61 1370 17000 10 113 2749 46 2749 29000 Table 1 Evaluation of the effectiveness of construction safety management information Since the order of data in the table is very large, to facilitate the solution of linear programming problems, standardization of the evaluation indicators, set up Zj = 1 10 10 10 Z ij , S j i 1 = 1 10 (Z i 1 ij Z j )2 ; Standardized indicators is Z Z ij' = Sijj The standard input and output indicators of table 1 is displayed in table 2 DMU Serial X1 X2 X3 Y1 Y2 1 2.8296 4.0664 5.5339 4.2466 3.6053 2 1.8392 2.6798 7.5104 3.0253 4.5667 3 1.4431 1.6968 5.0069 1.8608 2.4035 4 0.4668 0.9696 5.1387 1.3954 1.4421 5 0.6366 0.9090 6.5880 1.9780 1.3219 6 1.3582 2.5574 5.5339 3.3973 2.4035 7 3.9614 2.2423 6.4563 3.0253 3.6053 8 1.2011 1.4641 7.3786 1.0472 2.2833 9 0.9620 1.2120 8.0374 1.1635 2.0430 10 1.5746 3.3310 6.0610 2.3348 3.4851 numbers Table 2 Evaluation index of the standard value Applying DEA (C2R) model on standardized data ,the evaluation of the validity of the results obtained DEA in Table 3. 112 DMU Serial RMS Slack variable numbers 1 s1 s2 s3 s1 1 0 0 0 0 2 1 0 0 0 0 3 0.8881 0 0 0 0 4 1 0 0 0 0 5 1 0 0 0 0 6 1 0 0 0 0 7 1 0 0 0 0 8 0.9151 0.1796 0 2.9974 0.4654 9 0.9891 0.1288 0 4.5902 0.1898 10 0.9419 0 0.9805 0 0.0813 Table 3 DEA efficiency evaluation results From Table 2 we can know value of project 1,2,4,5,6,7 is 1, so at least they are weak DEA efficient, then its slack variable s , s are also 0, that prove DEA effective unit, indicating that the 6 safety management of construction projects is relatively effective. The 3,8,9,10 project DEA RMS are less than 1, and it is neither for the weak effective and not effective for the DEA. This number can indicate that the security management of these companies is ineffective, and accidents are occurring frequently in this way. Therefore, more attention should be paid to renovation. The smaller the Value is, the smaller the project safety management performance is, and the worse the management is. 5 Conclusion In this paper, a new method C2R DEA model is applied to evaluate the construction safety management performance evaluation . Through the example, the method is feasible and reliable. But the treatment of the two output variables is needed to deepen the study, to make it more objective and accurate. References [1]. Construction safety inspection standards JGJ 59-99(in Chinese) [2]. SHUANG Liang, ZHANG Shoujian. Construction Safety Fuzzy Comprehensive Evaluation [J]. Journal of Harbin Institute of Technology (Natural Science), 2003, (11):1357-1362 (in Chinese) [3]. SUN Shoupeng, YU ZHijun. Study on Construction Safety Evaluation [J]. Journal of Qingdao Technological University, 2009, 30(6): 45-52(in Chinese) [4]. WEI Songling .The Data Envelopment Analysis[M].Science Press, 2004: 1-20(in Chinese) [5]. FANG Dongping, HUANG Xinyu, Jimmie Hinze. Construction Safety Management [M]. Beijing: China WaterPower Press, 2003:73-124(in Chinese) 113

1/--pages