Improving planning reliability and project performance using the rational commitment model

The paper Improving planning reliability and project performance using the rational commitment model proposes the Rational Commitment Model (RCM), a new decision decisionmaking tool based on lean principles, which uses statistical models to obtain more reliable commitment planning at operational level. RCM allows forecasting commitment planning for short termperiods using information about workers, buffers, and planned progress. Đề tài Hoàn thiện công tác quản trị nhân sự tại Công ty TNHH Mộc Khải Tuyên được nghiên cứu nhằm giúp công ty TNHH Mộc Khải Tuyên làm rõ được thực trạng công tác quản trị nhân sự trong công ty như thế nào từ đó đề ra các giải pháp giúp công ty hoàn thiện công tác quản trị nhân sự tốt hơn trong thời gian tới.

m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u tv f5 rk sb 78 w9 p8 zư gh hg w m 9y s8 s0 Improving Planning Reliability and Project Performance using the Rational Commitment og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k Model g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw c7 sz Vicente González1, Luis Fernando Alarcón2, Sergio Maturana3, Fernando Mundaca4 and José Antonio c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 Bustamante5 2d om bp av d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 Abstract bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 Reliability of commitment planning at an operational level is one of the key factors for improving project ng wb v9 r6 db 1ư 7o jm cư fm hi 7c h lkx tư c8 2z ưb j 56 ai8 9d performance During the last fifteen years, the use of the Last Planner System (LPSTM), a production ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z vc iư ik6 lk 7s g4 2s 3u 4g 9h gy planning and control system based on lean production principles, has improved commitment planning 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c reliability in the construction industry However, commitment planning using the LPSTM has followed the p8 y5 z8 px vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u u j9s sz vo 3e 5y 3ư 6ư prevalent pattern in construction, in which decisions are mainly based on experience and intuition This 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m tj 07 m pt v 4o r 3lr j88 0r has limited its potential for improving commitment planning reliability To overcome these limitations, 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm this paper proposes the Rational Commitment Model (RCM), which is a new operational decision-making i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l pc kj tool based on lean principles RCM uses statistical models to develop more reliable planning vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa t fu 3l9 hz commitments at an operational level RCM allows forecasting commitment planning for short term8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u periods using common site information such as workers, buffers, and planned progress RCM has been pa kn ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt po 42 tested in several case studies, demonstrating its production forecasting capabilities and its ability to help 7l of ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw increase commitment planning reliability, thus improving project performance In addition, it is shown k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m how RCM allows balancing load and capacity Finally, the potential impact on the construction industry m j rk gw d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w b s 1ư icc s1 of RCM, as a simple and practical production decision-making tool, is briefly discussed jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 kư 9x 29 8q 0z pv Postdoctoral Fellow, Depto Ing Gestión de la Construcción, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile Assistant Professor, Escuela de Ingeniería de la Construcción, Universidad de Valparso, Chile (Corresponding Author) E-Mail: vagonzag@uc.cl Professor, Depto Ing Gestión de la Construcción, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile E-Mail: lalarcon@ing.puc.cl Professor, Depto Ing Industrial y de Sistemas, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile E-Mail: smaturan@ing.puc.cl Well Engineer, Integrated Project Management, Schlumberger E-Mail: mundaca@slb.com Planning and Development Engineer, Socovesa, Santiago, Chile E-Mail: jabustamante@socovesa.cl ied g ho yq kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv Keywords rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 f7 4p 5k 8p Lean Production, ts qg Decision-Making, 2i xd wz q m wu ge 6k Matching Load-Capacity, Rational Commitment Model, g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp lh fu Commitment Planning Reliability, Project Performance, Statistical Models 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb 66 Introduction b5 4y 28 2d om bp av d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư Planning has traditionally been a topic of much interest among academics and practitioners in project 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 management, due to its impact on performance during the execution phase In construction, however, the 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi focus of this attention has been on the development of planning tools rather than on theoretical issues 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp jh cu promoting its improvement (Laufer et al, 1994) In general, the non-recognition of theoretical issues in 6b 1z vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd construction management is a common pattern, leading to inadequately deal with production in projects kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 g ilw As a result, a poor project performance is achieved (Koskela, 2000) Otherwise, construction planning cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 sim b follows a similar behavior leading to a low performance as well (Ballard, 2000) fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 8b 6m Among the theoretical issues of construction planning, how planning decisions are made to manage 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge ar e6 variability in construction projects is one of the most relevant (Laufer et al, 1994) In general, variability d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 y ba isa 9r in construction depicts varying production rates, labor productivity, schedule control, cost control, etc hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b 7r h5 no 82 3f 3q j9 52 hm yc Variability is a well-known problem in construction on which there is much ongoing research (Ballard, eq y0 ag 5u pa kn ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 jrk u 1993; Tommelein et al, 1999; Thomas et al, 2002; among others) Some authors have recognized that 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 traditional project management does not consider the non-linear and dynamic nature of projects yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 hq le (Bertelsen, 2003 and McGray et al 2002) In construction, this yields non-realistic planning outputs (e.g wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w b schedule and budget), since planning process is based on the wrong notion that projects are static s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg Therefore, construction planning leads to poor management decisions since variability is not explicitly 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq incorporated within planning process, contributing to deteriorate project performance (González, 2008) kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu s3 11 Current construction planning not only follow the same pattern of traditional project management, where om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 the effect of variability impacts is systematically neglected, but it also depends on intuition and gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 experience to deal with variability As a result, planning has not been effectively managing projects and rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv has not been able to accurately predict how a project should be executed (Laufer et al, 1994), imposing rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 unrealistic expectations on the production process or failing to manage it altogether, increasing system f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 variability (Tommelein et al, 1999) 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc The Last Planner System (LPS™) (Ballard, 2000) was developed to overcome these limitations in 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av d fjy xv s7 jm 67 n5 y9 construction planning LPS™ is a production planning and control system based on lean production jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a principles Lean production is a management philosophy focused on adding value from raw materials to ưh m r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk finished product It allows avoiding, eliminating and/or decreasing waste from the value stream One of q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi these wastes, reducing variability, is the core of lean philosophy (Womack and Jones 1996) LPSTM 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z provides a management technique to deal with variable project behavior LPSTM promotes improved vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ftx tcd ck x ng planning reliability, which provides a stable production environment in projects and reduces the negative kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u impact of variability This creates reliable work plans monitored through the percentage of plan u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim f hq c8 m fx completed (PPC) in a short-term planning horizon (operational level) tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm Frequently construction projects outsource most of the work to subcontractors, and commitments are i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc arranged between contractors and subcontractors Contractors should strive to obtain reliable vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa t fu 3l9 hz commitments from the subcontractors However, many of them assign work to subcontractors based on 8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u their intuition and experience, resulting in unreliable commitments (Bustamante, 2007; Sacks and Harel, pa kn ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt po 42 2006) Although LPS™ represents a sounder planning framework, it does not deliver an entirely rational 7l of ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 7i 3p planning process mainly at operational level where the work is executed 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk This paper proposes the Rational Commitment Model (RCM), a new decision decision-making tool based m m xx 5y iư 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a on lean principles, which uses statistical models to obtain more reliable commitment planning at 3q vy a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re operational level RCM allows forecasting commitment planning for short term-periods using information bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m bp tfb qv about workers, buffers, and planned progress tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 04 yc The following sections in this paper describe the objectives, research methodology, and literature review, h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli o yk cf4 2a as well as the theoretical and practical foundations of the RCM Then, the validation process for the RCM 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv and its effect on planning reliability and project performance, and the load-capacity matching problem, in rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 several case studies are addressed f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp lh fu 7k 1c v6 o1 d2 Research Objectives lv fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om The main objective of the research was to develop an operational decision-making tool to improve bp av d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 reliability of commitment planning and project performance, by using statistical models and site u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t information of activities such as workers, buffers and planned progress, which we have called the yq re zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 Rational Commitment Model (RCM) db 1ư 7o jm cư fm hi 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z A secondary objective was to demonstrate how the RCM can help to match work load with on-site vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 tcd ck x ng ftx capacity, i.e., how it can help to balance the workload involved in work plans with the real labor capacity kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 ilw g of crews, by generating more reliable commitment planning for both contractors and subcontractors cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg vb 6o 1a qd ei xx sa yu Research Strategy and Methodology ylc b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln This paper uses the case study approach (Yin, 1994), supported with statistical analysis of data obtained fa ge e6 ar d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư from several construction sites This allowed us to validate the research assumptions and the RCM, as jz t oo 3l9 9r y ba isa hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 2b 35 well as to determine the impact on project performance of using RCM 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk The research methodology we use consisted of four stages: 1) Statement of theoretical foundations and 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 practical framework for the RCM; 2) Definition of case studies, where repetitive processes in multifamily yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq residential, multi-story building and industrial projects were studied; 3) RCM development and validation wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w process where multiple linear regression (MLR) techniques were used By using weekly on-site b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg information from case studies such as planned and actual workers, buffer sizes and planned and actual 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq progress at activity level, MLR models were constructed and validated A two-stage approach for the kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy ic validation process was applied, in which the first stage studied the robustness of mathematical 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 h8 w8 sd ld formulation of the RCM, and the second stage principally analyzed the RCM validation and application aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv in-action; and 4) Analysis and evaluation of RCM impacts on planning reliability and project performance rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 using several construction samples f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp lh fu 7k 1c v6 o1 d2 Last Planner Basics lv fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om av Overall Framework bp d fjy xv s7 jm 67 y9 n5 jz 92 4.1 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e LPS™ acts over the three project planning levels: (i) ‘Initial planning or master plan’ (strategic level), zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh which produces the initial project budget and schedule, and provides a co-ordinating map that ‘pushes’ tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi 7c h lkx tư c8 2z completions and deliveries onto the project (ii) ‘Lookahead planning’ (breakout of master plan – tactical ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z vc iư ik6 lk 7s level), which details and adjusts budgets and schedules ‘pulling’ resources into play (iii) ‘Commitment g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd kb op planning or work plans’ (short-term period – operational level), which regards the activities and schedule 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u j9s u work that will be done on-site according to the status of resources and prerequisites (Ballard and Howell sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc 1998) vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm 4u i8 The traditional management approach for work plans defines activities and schedule work that will be jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc vj 4y kc 8d done, in terms of what should be done from a master plan Crews are being committed by management to s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 ii8 8h v 4t whatever the schedule says should be done, with no real consideration for what they are actually able z7 7k bg z8 t2 a0 35 2b 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u kn pa to Then, waste such as idle time or ineffective work, among others, could affect crews due to ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no unforeseen variations of workflow In construction, workflow is characterized by crews moving from m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr location to location and completing the work that is prerequisite to starting work by the following crew c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m j rk gw m (Tommelein et al, 1999) In general, a reliable workflow depends on the following construction d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 y z8 1v preconditions that should be available whenever they are needed: resource (design, components and 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm materials, workers, equipment, space) and prerequisites (complete work of upstream activities) (Koskela, q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq kp x1 gt 80 dn ưi 2000) ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 7w rx LPS™ overcomes previous limitations providing a predictable production environment in projects, 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg tli z5 fo decreasing workflow variability and creating reliable work plans to derive maximum project benefits The 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv overarching criterion in the LPS™ is that activities should only be committed if they can be performed rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 (i.e all resources and prerequisites that are needed must be available), transforming what should be done f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 into what can be done, from which a work plan can be formulated Thus, work plans will be based on 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl achievable assignments serving as a commitment to what will actually be done The following criteria are be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av critical for the assignments: 1) they are well defined, 2) the right sequence is selected, 3) the right amount d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư z a3 i47 8q of work is selected, and 4) the work selected is practical or sound, that is, can be done according to the u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t re yq availability of construction preconditions, which are considered during the lookahead planning zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi 7c lkx h In this paper, the notion of reliability is focused on project planning, so that PPC depicts a project tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z vc iư ik6 planning reliability index PPC is understood as the ratio between actual completed activities and planned lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ftx tcd ck x ng activities A low PPC means unreliable planning and a high PPC near 100%, means the opposite From a kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u Lean Production perspective, PPC is also a measure of workflow variability A low PPC is an indicator of u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim f hq c8 m fx a highly variable workflow as a result of unfavourable project conditions to execute planned activities A tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 high PPC indicates stable and predictable workflow as a result of favourable project conditions to 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh j jj8 q7 8m fln ob c bb complete planned activities fa ge e6 ar d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 v 4t ii8 Relationship between Planning Reliability and Project Performance 8h z7 7k bg z8 t2 a0 35 2b 7r h5 4.2 no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn ro u5 l9q 2ig 93 wd p2 o5 c6 71 LPS™ has been applied in numerous projects around the world in the last fifteen years, being reported b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns 6v jq performance improvements on a wide range of construction projects (Alarcon et al 2005, Liu and Ballard ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e 2008, González et al, 2008a, among others) The main assumption of the system is that an increase in dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx iư 5y planning reliability, measured through PPC, should improve project performance, and mainly, 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk productivity Recently, several researchers have demonstrated a positive and strong relationship between m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 9x kư planning reliability and project performance, where the impact of a better planning reliability has been 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx measured through the improvements over productivity at project level (González et al, 2008a; Liu and vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 Ballard, 2008) yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv Due to the limited evidence linking the changes in planning reliability with changes in productivity at the rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 activity level, an in-detail study was carried out by González et al (2008a) Sacks and Harel (2008)´ f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 research addressed a similar issue, but only from a theoretical viewpoint These authors proposed 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl reformulation of the indicator for planning reliability used by LPS™, PPC, to carry out meaningful be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av productivity comparisons at the activity level Therefore, a complementary ‘activity-based’ planning d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư z a3 i47 8q reliability index, called Process Reliability Index (PRI) was developed PRI is defined as (González, et al, u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 2008): ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z   × 100   vc iư ik6  APi , j PRI i , j =   PP  i, j (1) lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 where: g ilw cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim PRIi,j= Process Reliability Index for week i and activity j (%), i=1…n; j=1 m fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 1o 2i APi,j= Actual Progress for week i and activity j, i=1…n; j=1 m 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc vj 4y kc 8d PPi,j= Planned Progress for week i and activity j, i=1…n; j=1 m s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b 7r h5 PRI represents a planning reliability index at the activity level PRI does not compare actual to planned no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn ro u5 l9q 2ig 93 wd p2 o5 cumulative progress because it is based on partial measurements (i.e weekly progress), which can vary c6 71 b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns jq 6v from a measurement period to another PRI measures the degree of activity planning effectiveness from a ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc lp 3o commitment standpoint To measure planning reliability, PRI values range between and 100% When t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m xx m AP is higher than PP, the PRI value is limited to 100% (González et al, 2008a) 5y iư 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg A study of the relation between PRI and productivity at activity level by González et al (2008a) showed 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq that higher PRI levels lead to improved productivity This confirms the assumption that increasing kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy planning reliability improves project performance at activity and project level In this sense, LPSTM acts at ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 h8 w8 ld sd project level, producing planning reliability improvements not only at that level, but also at activity level aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp qv e4 to get improvements in a project as a whole ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 Matching Load and Capacity tv 4.3 rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 f7 4p 5k 8p ts qg 2i xd wz q m Matching load with capacity is critical for productivity of production systems in construction (Ballard, wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp lh fu 7k 1c v6 o1 lv d2 2000; Hassanein and Mellin, 1997; Thomas and Horman; 2006, among others) According to Ballard fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 (2000), load is the amount of work in a specified time which is assigned through planning to crews In 8m 3e zx kb b5 66 4y 28 2d om bp av d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 contrast, capacity is the amount of work a crew can at any point in time with given tools and work p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a ưh m methods for actual site conditions The problem in matching load with capacity is that, for instance, actual r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk q4 resource utilization and production rates of crews are production variables many times a-priori unknown, dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi 7c h lkx c8 tư given their changeable behavior caused by wastes in conventional practices (Ballard, 2000), leading to a 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z iư ik6 vc poor balance between load and capacity and losses of productivity lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 z8 px Ballard (2000) states whatever the precision of load and capacity estimates, the planner must still make vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 some adjustments in which load can be changed to match capacity by delaying or accelerating workflow, xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 ylc 3t r sg capacity can be changed to match load by decreasing or increasing resources, or more commonly, a vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm combination of the two However, the preference seems to be for adjusting load LPSTM is instrumental to i8 4u jt gv g0 8m fq k gs rrh ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc 4y vj match load with capacity by pulling materials and/or information into a production process or activity, kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 8h v 4t ii8 only if the activity is able of doing the work, i.e., what activity needs and in the needed amounts are z7 7k bg z8 t2 a0 35 2b 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn actually available The pulling process is carried out within Look Ahead planning (see Womack and ro u5 l9q 2ig 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt 42 po of 7l Jones, (1996) or Hopp and Spearman (2000) for more details about pull systems) ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 3p 7i However, LPSTM can loss effectiveness to match load with capacity from the lookahead planning to the 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m work plans This issue can emerge when critical criteria for work assignments are not correctly defined m xx 5y iư 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy and met during the lookahead planning Several reasons can explain it First, it is difficult to accurately a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re xl4 bj a ac determine the right amount of work to perform by crews in work plans based only on the experience of kư 9x 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m qv bp tfb eh tb project personnel since it can be not very reliable and subjected to several biases (Spetzler and Von zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w Holstein, 1975, McGray, et al, 2002) In contrast, if historical data is used, it may not be accurate enough, 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp due to changes in current construction practices (Ramírez et al, 2004) e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 8 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv Second, it is not easy to see if work is practical or sound (i.e all construction preconditions are ready for rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 crews to perform work) for work plans For instance, the number of necessary site-workers supplied by f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 subcontractors to a specific project depends on business demands, i.e labor requirements from other 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl projects If there are projects with better site conditions where the subcontractor job is more profitable, his be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av preference will be change the labor resource until site conditions are improved in the original project d fjy xv s7 jm 67 y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư z a3 i47 8q Thus, labor resource can constantly be changed from one project to another (in a weekly or even daily u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t re yq basis) (Lee et al, 2004; Sacks and Harel, 2006) As a result, labor resources are not ready or available zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư whenever it is required and in the correct amount in a project Another example is when the work of a 7o jm cư fm hi 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 crew in repetitive processes is dependent on the activity progress of an upstream crew This is also known 6lp cu jh 6b 1z vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs as work-in-process buffer, which characterizes work units ahead of the crew that are necessary to perform 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve its own work (González et al, 2009) Many times these buffers are not available whenever the crew needs vb gl s8 h3 g ilw cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c them, since their own rate of production can be unpredictable (even at a daily level) (González, 2008; 7s 34 b sim fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o qd 1a Tommelein et al, 1999) Therefore, this precondition will be not available in the necessary timing ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh 8m fln ob c bb j jj8 q7 These issues address several limitations related to how to match load with capacity in the LPSTM, fa ge e6 ar d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu f6 3z o2 j fzư jz t oo 3l9 suggesting to change the way in which this process is carried out In this paper we argue that more 9r y ba isa hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b 7r h5 82 no rational mechanisms could support a better load and capacity matching process, thus improving the 3f 3q j9 52 yc hm eq y0 ag 5u pa kn ro u5 l9q 2ig 93 wd p2 o5 71 c6 quality and reliability of work plans b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui Intuition and Rationality for Making Commitments Planning lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx iư 5y Most of the people tend to describe and understand the world around through simplistic models of reality 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk This may be due to the difficulties that human beings have to manipulate large amount of information, m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 9x kư developing in many cases mental twirls (Spetzler and Von Holstein, 1975) In construction, this kind of 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx phenomena is prevalent in its decision-making processes given the complexity and dynamic nature of the vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 projects, which can lead to erroneous and poor decisions (Bertelesen, 2003; McGray et al, 2002) For yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg fo tli z5 instance, a common practice for estimating labor productivity, and accordingly, construction schedules 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv and budget, is to simply assume that work progress in projects is very stable Fig shows an illustration rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 of this behavior by using simple linear regression models in a hypothetical construction activity In these f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 0f py pe jp models, for every x-value there is an expected y-value ( y ) with a certain standard deviation (σ) (y is an lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t dl be pi d1 t0 kư aleatory variable which follows a normal distribution) The regression model in Fig 1.a represents a 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av d fjy xv s7 67 jm typical estimation of labor productivity by using an 'average estimated' based on historical data, where the y9 n5 jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 r z9 tjp u5 coefficient of determination (R2) equal to 1.0 expresses a perfect fit to data of regression model This is an bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ideal and non-realistic situation, since it suggests a deterministic behavior for labor productivity on-site ưk q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi In contrast, Fig 1.b shows the same issue but regarding an intrinsically 'variable estimated' of labor 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z productivity, which is captured using a regression model where site behavior of labor productivity is vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl tư qs 7d ju 1m 86 modeled for a more coherent R2-value equal to 0.85 Also, this model characterizes an imperfect ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 z8 px vx m o1 ve vb gl s8 h3 ilw g relationship between production rates (probabilistically described by an expected value plus a σ), and the cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m number of workers (held constant), although more fitted to a real production situation tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m fq k gs rrh fln b) Production rate v/s Worker-days ob c bb j jj8 q7 8m a) Production rate v/s Worker-days fa ge e6 ar d l3l P ro babilistic Value y=12.8 (wo rk units/day) σ=1,18 (wo rk units/day) kj pc vj 4y 14.0 jz t oo 3l9 9r hz t fu 3l9 12.8 8h v 4t ii8 12.0 y ba isa z7 7k bg z8 10.0 t2 a0 35 2b 7r h5 no 82 3f 3q 8.0 j9 52 yc hm eq y0 ag 5u kn 6.0 pa ro u5 l9q 2ig 93 wd 4.0 o5 y = 1,89x + 3,33 R2 = 0.85 p2 c6 71 b l1ư vw 17 u jrk 2.0 9s pt 42 po 7l of ib no m c1 0.0 sn zd iao 9d ns y - Production rate (work units) o2 0.0 j 2.0 fzư 4.0 f6 y = 2.80x R2 = 3z 6.0 47 8.0 nu 10.0 q 12.0 irc 14.0 16.0 s8 Deterministic Value hp y - Production rate (work units) kc 8d 16.0 jq 6v 6.0 0.0 1.0 2.0 3.0 xs 5.0 94 4.0 69 3.0 uj 2.0 w7 1.0 ds 0.0 4.0 5.0 6.0 vư c7 yy vv dw x - Worker-days 59 x - Worker-days (wd) k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx 5y iư 9y xq Fig Labor Productivity Estimates: (a) Deterministic Approach; (b) Probabilistic Approach oi y3 ưz 3h i4w b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 To illustrate the idea we will compute the productivity for a given number of workers A closer analysis kư 9x 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m bp tfb qv of the information provided by the Fig 1.a and Fig 1.b shows that for worker-days a production rate for tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 the crew of 14.0 (work units/day) and 12.8 (work units/day) can be expected respectively Otherwise, Fig rx 7w 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 vx gt 1.b provides probabilistic information for both estimates In this figure, production rates follow a normal 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily ưz k tu 67 q8 ji 43 1r rb distribution with a σ equal to 1.18 (work units/day) Then, cumulative probabilities to achieve an equal or wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 10 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv predicted PRI of 100% Table shows that the level of W does not change for points and Finally, rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 a project manager will select a predicted PRI value and a production frame from the sensitivity f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 analysis according to his preferences (e.g labor cost for a higher W level, time to produce a higher 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t WIPBf size, among others) dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av d fjy xv s7 jm 67 n5 y9 8) Definition of the planned progress: By using production frame from stage or 7, a planned progress jz 92 yg y7 p6 e0 tvt 51 5g 6k ux ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a is estimated at the beginning of a labor week Two decisions can be taken: i) RCM prediction is used ưh m r0 2e zh 42 lo p2 bh gb ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk as planned progress to actual week, or ii) Project managers keep their own estimate as planned q4 dh tư 1u e7 wi hw d0 78 ng wb v9 r6 db 1ư 7o jm cư fm hi 7c lkx h progress (the common process would gradually be to change from manager estimate to RCM tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 0n em 7j9 vư e 60 6lp cu jh 6b 1z vc iư ik6 lk 7s g4 2s 3u 4g 9h gy estimate) 35 2a z lvy kjl tư qs 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 9) Final data collection: The final data is gathered at the end of labor week which is necessary to z8 px vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u u j9s sz vo 3e 5y 3ư 6ư further evaluation of RCM predictions The data measured are AP and actual W 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx 10) Evaluation of the RCM predictions: Once a labor week has finished and regarding the necessary sa yu b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m rrh fq k gs data, the main accuracy measures for the RCM prediction are computed, that is, predicted and actual ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu PRI, and predicted and planned CCL This is a key stage to evaluate the quality of RCM predictions f6 3z o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 Finally, the RCM process is repeated in Stage 11 (similar to stage without AP and actual W 35 2b 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn u5 l9q 2ig ro information) and stage 13 (similar to stage without intermediate decisions) until the activity has 93 wd p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no m c1 iao sn zd been completely executed 9d ns jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp t4 a0 k8 2e k1 RCM Role for Improving Planning Reliability and Project Performance dl fc 16 le hq wv wg m m j rk gw 7.1 d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w b s1 s 1ư icc jy9 y z8 1v The use of statistical models in the RCM to describe the production behavior in projects will inherently 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 54 kw 9w hg 7g wu 51 x0 gm 88 increase planning reliability at activity level On the other hand, by means of increasing planning q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g ho yq kp x1 gt 80 ưi dn ld 9m reliability is possible to improve performance in projects at two levels: project and activity Therefore, qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy ic 12 67 nb 38 e4 fp da cu 11 s3 improvements on planning reliability and performance (labor productivity) at activity level will lead to om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 h8 w8 sd ld aq pc u1 6y oi 3ư yu 4r enhance the same thing at project level In fact, common sense suggests that if a set of activities p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 4d rp e4 qv ưv vz lw v6 ily k tu ưz individually increase its planning reliability and this set structures the entire project, then it is expected 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 19 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k m qm g2 zs bc x4 uư wh od wg q3 ưj 1w 1n 0d 4v nq c3 bg y5 k2 ưl 4p g fl2 2x 77 0x 9y z4 r9 3f m 3o c hy 7s sc 8g d2 Journal of Construction Engineering and Management, ASCE (submitted on April, 2009) 3a r nv yjv uf d6 y5 gv 7l3 ba g6 o9 74 n1 yc 0q o3 3n t0 nk jl cg wd qv va nh ql sn wm ql 5o ffr u f5 tv that the planning reliability at project level is improved getting a better performance at project level rk sb 78 w9 p8 zư gh hg w m 9y s0 s8 og lb nj qt 4iu df ư5 u0 7q jm rb y4 r4 Previously, empirical evidence and analysis performed by González et al (2008a) supported this notion f7 4p 5k 8p ts qg 2i xd wz q m wu ge 6k g2 20 ux 7k zp jeu pe 5x xk k6 As result, RCM action to improve planning reliability and project performance starts at activity level to 0f py pe jp lh fu 7k 1c v6 o1 lv d2 fo 48 17 83 kw sz c7 c uk 3lv oe hd 2t finally act at project level dl be pi d1 kư t0 3s 7ư trk jcc 2c j z9 8m 3e zx kb b5 66 4y 28 2d om bp av d fjy xv s7 jm 67 y9 n5 Matching Load and Capacity with the RCM jz 92 yg y7 p6 e0 tvt 51 5g 6k ux 7.2 ln uư 8q z a3 i47 u5 r z9 tjp bl dư 6s oj 0z 0a ưh m r0 2e zh 42 lo p2 bh gb The capability to match load with capacity is another characteristic of the RCM Two mechanisms are ku 2t yq re zj hd f2 3d ui 96 2k i vm jc5 ưk q4 dh tư 1u e7 wi hw d0 78 ng wb basically applied by the RCM to match load with capacity: 1) Fix either load or capacity and develop v9 r6 db 1ư 7o jm cư fm hi 7c h lkx tư c8 2z ưb 9d j 56 ai8 ws kp b kj6 sensitive analyses for the free variable according to actual production conditions, and 2) Study the effect 0n em 7j9 vư e 60 6lp cu jh 6b 1z vc iư ik6 lk 7s g4 2s 3u 4g 9h gy 35 2a z lvy kjl of several construction preconditions that can prevent the performance of an activity to mitigate its tư qs 7d ju 1m 86 ck x ng ftx tcd kb op 6x dq eu 2c p8 y5 z8 px impact To understand both mechanisms Fig will be used vx m o1 ve vb gl s8 h3 g ilw cz jp 7d 1u u j9s sz vo 3e 5y 6ư 3ư 8j9 xt 1a 30 jb ưh 8c 7s 34 b sim In the first mechanism, load as planned progress in Fig can be fixed analyzing the level of capacity as fx f hq c8 m tj 07 m pt 0r v 4o r 3lr j88 3t r sg ylc vb 6o 1a qd ei xx yu sa worker-weeks required to meet the amount of work planned Other variable involved in the estimation of b0 z qr m ư0 xo a1 2i 1o 6m 8b 1ư rp qn tm i8 4u jt gv g0 8m k gs rrh fq load is the planned PRI which a decision-maker can include to visualize the impact of the planning ob c bb j jj8 q7 8m fln fa ge e6 ar d l3l kj pc vj 4y kc 8d s8 hp q irc 47 nu f6 3z reliability over capacity levels In contrast, if capacity is limited, i.e the number of worker-weeks is o2 j fzư jz t oo 3l9 9r y ba isa hz t fu 3l9 8h v 4t ii8 z7 7k bg z8 t2 a0 35 2b constrained by the subcontractor’s needs; the level of load is adjusted to certain amount of work given a 7r h5 no 82 3f 3q j9 52 yc hm eq y0 ag 5u pa kn ro u5 l9q 2ig wd 93 planned PRI Obviously, it can be a third option in which both load and capacity can be simultaneously p2 o5 c6 71 b l1ư vw 17 u jrk 9s pt 42 po 7l of ib no m c1 sn zd iao 9d ns matched according to the decision-makers preferences Also, the extent for which both load and capacity jq 6v ds w7 69 uj xs 94 vư c7 yy vv 59 dw k6 ui lt pr c3 ho 7i 3p 2b qc 3o lp can change week to week is determined by the information statically processed in the RCM t4 a0 k8 2e dl k1 fc 16 le hq wv wg m m j rk gw d0 cn 40 0f 1a xk m m xx 5y iư 9y xq oi y3 ưz 3h i4w In the second mechanism, RCM explicitly manipulates several construction preconditions as number of b s1 s 1ư icc jy9 1v y z8 9t xr hz 4n x1 1a 3q vy a2 kk m cư zk a h9 kw 54 workers and buffer levels The first precondition is analyzed according to the first mechanism Otherwise, 9w hg 7g wu 51 x0 88 gm q7 09 r9 re bj a ac xl4 kư 9x 29 8q pv 0z ied g the analysis of buffer levels is one the most interesting characteristics of the RCM previously studied by ho yq kp x1 gt 80 ưi dn ld 9m qv bp tfb tb eh zd 8c y7 f fu ffx vm 1o oy González et al (2008b) For instance, the influence of the buffer size (WIP Bf) over labor productivity ic 12 67 nb 38 e4 fp da cu 11 s3 om 1c 8y v5 rx 7w 5a zu 1c e6 yc 04 w8 h8 given the planned progress (load) and planned PRI can be analyzed Fig shows this behavior A larger sd ld aq pc u1 6y oi 3ư yu 4r p2 b1 gt vx 9s xg z5 fo tli 2a o yk cf4 rp 4d buffer results improved labor productivity, and therefore in a reduced number of worker-weeks On the e4 qv ưv vz lw v6 ily ưz k tu 67 q8 rb ji 43 1r wa gm li t2 1x c2 ki lp 5d 70 ys fl ib xf g0 62 wg 73 bl bt i6 2x g1 ue 0ư b1 ua m x eq m 55 t 8f ta 9c 5y 9w dư kr 3m 6ư f 8q m ri ưm 6m lcư 20 0iq 4w pw r 3n 5c ưk 23 ef r7 df d9 c uv m 50 ux a7 iv n9 ym jki bl j 7o x4 73 5h f0 6q be n0 gd we k