Progress in Large-Shared Projects
Content
Progress in Large-Shared projects: Method for forecasting and optimizing project duration in a distributed project
LGC - PSI - Département <<Génie Industriel>>, INPT-ENSIACET, 118 route de Narbonne, 31077 Toulouse cedex 4, France
Rameez KHALID, Alain HAIT, Philippe DUQUENNE
Plan
Introduction Literature Overview Problem Statement Simulation Results Conclusion
1
Introduction
Large-Shared projects:
Large in respect of: No. of decisions to make, tasks, participants, significant cost, information size, and substantial risks; Shared such that: they depend on more than one entity for their completion and the project (or major parts of it) can be indifferently distributed to one or another resource.
Transferable Work-load Project:
a project in which work is transferable from one resource to another even if job processing has started
OBJECTIVE: optimization of project duration with the help of progress measurement and redistribution of remaining work.
Literature Overview
The references of Project management such as IPMA, PMI and AFITEP are general. Every project is unique at some level, which requires specific methods for its problems. Every process of the project life-cycle requires proper attention in managing large-shared or transferable work-load projects. PROJECT PLANNING
It is an iterative process of developing a preordained plan in a forecasted environment. Planning can be strategic, tactical and operational. Planning requires: SOW, specifications, WBS, networks and schedules.
2
Literature Overview
MONITORING AND CONTROL
Earned Value Management (EVM):
Schedule Variance (SV=EV-PV) and Cost Variance (CV=EV-AV).
Progress Plot [12] :
Time on horizontal axis & progress on vertical axis with control lines.
Milestone and Resource slip charts [13] :
Uses a No. of review stages at which re-estimations can be done.
Performance indicators:
Model for performance evaluation of R&D projects proposed [14]. PMS created for improving supply chain process of Nike in [15].
FORECASTING
It is a process of previewing what will happen by certain time. For predicting the future precisely, accurate knowledge of system and its environment is indispensable. EVM is an early warning system which alerts during progress.
Problem Statement
Problem is related to the Pharmaceutical industry. Determination of the effectiveness of a treatment. Patients will be divided among a No. of countries, considering their population, capacity & competence. Two stages: Recruitment and Treatment of patients.
Three phases of treatment: each phase is of a duration, a certain part of treatment and a different defined cost. € Treatment Phases
1 2 2nd payment 1st payment Time 3
3
Problem Statement
Scheduled budget is calculated on the basis of recruitment planning. Countries will be paid at every update depending upon the treatment phase completed by each patient. As planning ends, actual recruitment data is generated. OBJECTIVE: to analyze how the system responds in different situations and thus optimizing project duration.
Problem Statement
PLANNING AND BUDGETING
Three periods in the execution of a project in S-Curves, namely: learning, working and ending periods. Parkinson’s Law i.e. “Work expands to fill the time available”. The point where the learning period ends and the working period start is known as “Point of Inflexion (POI)”.
No. of
Patients
Variables: POI as a pre-specified no. of patients or duration The recruitment rate (d) A number of patients will be lost or will leave the system before the treatment ends.
Learning
Working
Ending
Time
4
Problem Statement
MONITORING AND CONTROL
Actual recruitment data are produced depending upon the performance (P). Two approaches to this problem, namely: with and without-reallocation. Linear programming is used for the reallocation with minimizing the total reestimated duration as objective. For the determination of “When the real POI will occur in time?” along with the rate of recruitment in the working period, a “Linear Hypothesis” is used.
Is
where,
IR I
d R1 d1
d R2 d2
Number of Patients N
N1+N2 = N1 +N2
Re-Planned
Schedule Performance Index = IS Planned learning period ends = I Real learning period ends = IR Recruitment Rate for previewed learning period = d1 Recruitment Rate for the real learning period = dR1
Country 1 N1 N1 N2 N2 Country 2 Time t t Tf1 Tf Tf2
Simulation
PROGRAM
Program is built in Microsoft Excel with programming in Visual Basic. If the file of pharmaceutical problem is opened, Form1 will be shown. The button “Solve” if pressed will generate scheduled budget and then Form2 will come up. The button “Generate Real Data” will generate real recruitment data. The frame “Point of inflexion: Achieved” provides a No. of options to continue this real data generation till the next update.
5
Results
ANALYSIS AND RESULTS
The simplest option available for each variable was chosen. This problem, involving numerous variables, offers interesting insight for managing transferable work-load projects. No. of instances are created by changing the values of variables. For every situation first results are gathered for withoutreallocation and then with-reallocation. With-reallocation is the champion approach for optimizing project duration in transferable workload or shared environment.
Performance entered for Country 1 & 3 90% & 75% Country 2, 4 & 5 110, 120 & 70% WithoutReallocation Duration 730 days Duration 1551 days WithReallocation Duration 669 days Duration 539 days 547 days Planned Duration
Results
ANALYSIS AND RESULTS
We are also interested in finding out “When to use reallocation?”. Reallocation is applied on the 1st update after POI and results were noted then on the 2nd update and so on. Transfer costs are also added for the transferred patients. Cost gets closer to 100% i.e. planned cost as we apply reallocation in later updates. Duration also gets close to what was planned.
102
% of Budgeted Cost
101.8 101.6 101.4 101.2 101 100.8 100.6 100.4 100.2 550 600 650 700 750 800 850 900
Actual Duration
6
Results & Extension
ANALYSIS AND RESULTS
Objective is to study how the system responds in different situations and thus optimizing project duration. With-reallocation is an effective approach. Reallocation used when 2/3 of the project duration passed will produce better compromise between duration and cost.
EXTENSION
Several avenues as an extension to this problem:
a penalty for late project completion; cost of adding a new country; transfer costs for transferring patients to another or a new country; these costs can be fixed or with a different cost coefficient
Conclusion
Not much attention has been given to the part of planning for transferable work-load projects, in literature. For better controlling and monitoring a project’s progress, techniques are still required that expose our planning flaws. Lessons learned may not be generalized but may be helpful in the realization of similar goals. Generalized propositions for further study may be:
improving risk management techniques and their applications describing project specific management methods.
7
Acknowledgment
Thank you for your attention
Questions
[1] Project Management Institute, A guide to the project management body of knowledge: PMBOK Guide, 3rd ed. Newtown Square, PA: Project Management Institute, 2004. [2] P. C. van Fenema and K. Kumar, "Coupling, interdependence and control in global projects," in Projects as Business Constituents and Guiding Motives, R. A. Lundin and F. Hartman, Eds. Boston, MA: Kluwer Academic Publishers, 2000. [3] B. Egginton, "Multi-national consortium based projects: Improving the process," International Journal of Project Management, vol. 14, pp. 169-172, 1996. [4] G. Bachy and A.-P. Hameri, "What to be implemented at the early stage of a large-scale project," International Journal of Project Management, vol. 15, pp. 211-218 1997. [5] R. A. Sprague, K. J. Singh, and R. T. Wood, "Concurrent engineering in product development," IEEE Design & Test of Computers, 1991. [6] H. C. Zhang and D. Zhang, "Concurrent engineering: an overview from manufacturing engineering perspectives,“ Concurrent Engineering: Research and Applications vol. 3, pp. 221- 236, September 1995. [7] G. Probert, "Projects, people and practices," Engineering Management Journal pp. 141-6, June 1997. [8] R. P. Smith, "The historical roots of concurrent engineering fundamentals," IEEE Transaction on Engineering Management, vol. 44, Feb 1997. [9] H. Kerzner, Project management – a systems approach to planning scheduling and controlling, 8th ed. New York: John Wiley, 1998. [10] Wysocki R and R. McGary, Effective project management, 3rd ed. New York: John Wiley, 2003. [11] N. Hardie, "The prediction and control of project duration: A recursive model," International Journal of Project Management, vol. 19, pp. 401-409, 2001. [12] M. J. Schmidt, "Schedule monitoring of engineering projects," IEEE Transaction on Engineering Management, vol. 35, May 1988. [13] B. Elphick, "Linking the use of resources to project progress," Engineering Management Journal pp. 279-282, 1992. [14] A. S. Pillai, A. Joshi, and K. S. Rao, "Performance measurement of R&D projects in a multi-project, concurrent engineering environment," International Journal of Project Management, vol. 20, pp. 165-177, 2002. [15] C. Lohman, L. Fortuin, and M. Wouters, "Designing a performance measurement system: A case study,"European Journal of Operational Research vol. 156, pp.267-286, 2004. [16] M. Bourgault, E. Lefebvre, L. A. Lefebvre, R. Pellerin, and E. Elia, "Discussion of metrics for distributed project management: Preliminary findings," in Proceedings of the 35th Hawaii International Conference on System Sciences, Hawaii, 2002. [17] D. S. D. Jacob and D. Associates, "Forecasting project schedule completion with earned value metrics," The Measurable News 2003, vol. 1, pp. 7-9, March 2003. [18] F. S. F. Patrick, "Getting out from between Parkinson's rock and Murphy's hard place," PMNetwork, vol. 13, pp. 49-51, 1999.
Bibliography
8
LGC - PSI - Département <<Génie Industriel>>, INPT-ENSIACET, 118 route de Narbonne, 31077 Toulouse cedex 4, France
Rameez KHALID, Alain HAIT, Philippe DUQUENNE
Plan
Introduction Literature Overview Problem Statement Simulation Results Conclusion
1
Introduction
Large-Shared projects:
Large in respect of: No. of decisions to make, tasks, participants, significant cost, information size, and substantial risks; Shared such that: they depend on more than one entity for their completion and the project (or major parts of it) can be indifferently distributed to one or another resource.
Transferable Work-load Project:
a project in which work is transferable from one resource to another even if job processing has started
OBJECTIVE: optimization of project duration with the help of progress measurement and redistribution of remaining work.
Literature Overview
The references of Project management such as IPMA, PMI and AFITEP are general. Every project is unique at some level, which requires specific methods for its problems. Every process of the project life-cycle requires proper attention in managing large-shared or transferable work-load projects. PROJECT PLANNING
It is an iterative process of developing a preordained plan in a forecasted environment. Planning can be strategic, tactical and operational. Planning requires: SOW, specifications, WBS, networks and schedules.
2
Literature Overview
MONITORING AND CONTROL
Earned Value Management (EVM):
Schedule Variance (SV=EV-PV) and Cost Variance (CV=EV-AV).
Progress Plot [12] :
Time on horizontal axis & progress on vertical axis with control lines.
Milestone and Resource slip charts [13] :
Uses a No. of review stages at which re-estimations can be done.
Performance indicators:
Model for performance evaluation of R&D projects proposed [14]. PMS created for improving supply chain process of Nike in [15].
FORECASTING
It is a process of previewing what will happen by certain time. For predicting the future precisely, accurate knowledge of system and its environment is indispensable. EVM is an early warning system which alerts during progress.
Problem Statement
Problem is related to the Pharmaceutical industry. Determination of the effectiveness of a treatment. Patients will be divided among a No. of countries, considering their population, capacity & competence. Two stages: Recruitment and Treatment of patients.
Three phases of treatment: each phase is of a duration, a certain part of treatment and a different defined cost. € Treatment Phases
1 2 2nd payment 1st payment Time 3
3
Problem Statement
Scheduled budget is calculated on the basis of recruitment planning. Countries will be paid at every update depending upon the treatment phase completed by each patient. As planning ends, actual recruitment data is generated. OBJECTIVE: to analyze how the system responds in different situations and thus optimizing project duration.
Problem Statement
PLANNING AND BUDGETING
Three periods in the execution of a project in S-Curves, namely: learning, working and ending periods. Parkinson’s Law i.e. “Work expands to fill the time available”. The point where the learning period ends and the working period start is known as “Point of Inflexion (POI)”.
No. of
Patients
Variables: POI as a pre-specified no. of patients or duration The recruitment rate (d) A number of patients will be lost or will leave the system before the treatment ends.
Learning
Working
Ending
Time
4
Problem Statement
MONITORING AND CONTROL
Actual recruitment data are produced depending upon the performance (P). Two approaches to this problem, namely: with and without-reallocation. Linear programming is used for the reallocation with minimizing the total reestimated duration as objective. For the determination of “When the real POI will occur in time?” along with the rate of recruitment in the working period, a “Linear Hypothesis” is used.
Is
where,
IR I
d R1 d1
d R2 d2
Number of Patients N
N1+N2 = N1 +N2
Re-Planned
Schedule Performance Index = IS Planned learning period ends = I Real learning period ends = IR Recruitment Rate for previewed learning period = d1 Recruitment Rate for the real learning period = dR1
Country 1 N1 N1 N2 N2 Country 2 Time t t Tf1 Tf Tf2
Simulation
PROGRAM
Program is built in Microsoft Excel with programming in Visual Basic. If the file of pharmaceutical problem is opened, Form1 will be shown. The button “Solve” if pressed will generate scheduled budget and then Form2 will come up. The button “Generate Real Data” will generate real recruitment data. The frame “Point of inflexion: Achieved” provides a No. of options to continue this real data generation till the next update.
5
Results
ANALYSIS AND RESULTS
The simplest option available for each variable was chosen. This problem, involving numerous variables, offers interesting insight for managing transferable work-load projects. No. of instances are created by changing the values of variables. For every situation first results are gathered for withoutreallocation and then with-reallocation. With-reallocation is the champion approach for optimizing project duration in transferable workload or shared environment.
Performance entered for Country 1 & 3 90% & 75% Country 2, 4 & 5 110, 120 & 70% WithoutReallocation Duration 730 days Duration 1551 days WithReallocation Duration 669 days Duration 539 days 547 days Planned Duration
Results
ANALYSIS AND RESULTS
We are also interested in finding out “When to use reallocation?”. Reallocation is applied on the 1st update after POI and results were noted then on the 2nd update and so on. Transfer costs are also added for the transferred patients. Cost gets closer to 100% i.e. planned cost as we apply reallocation in later updates. Duration also gets close to what was planned.
102
% of Budgeted Cost
101.8 101.6 101.4 101.2 101 100.8 100.6 100.4 100.2 550 600 650 700 750 800 850 900
Actual Duration
6
Results & Extension
ANALYSIS AND RESULTS
Objective is to study how the system responds in different situations and thus optimizing project duration. With-reallocation is an effective approach. Reallocation used when 2/3 of the project duration passed will produce better compromise between duration and cost.
EXTENSION
Several avenues as an extension to this problem:
a penalty for late project completion; cost of adding a new country; transfer costs for transferring patients to another or a new country; these costs can be fixed or with a different cost coefficient
Conclusion
Not much attention has been given to the part of planning for transferable work-load projects, in literature. For better controlling and monitoring a project’s progress, techniques are still required that expose our planning flaws. Lessons learned may not be generalized but may be helpful in the realization of similar goals. Generalized propositions for further study may be:
improving risk management techniques and their applications describing project specific management methods.
7
Acknowledgment
Thank you for your attention
Questions
[1] Project Management Institute, A guide to the project management body of knowledge: PMBOK Guide, 3rd ed. Newtown Square, PA: Project Management Institute, 2004. [2] P. C. van Fenema and K. Kumar, "Coupling, interdependence and control in global projects," in Projects as Business Constituents and Guiding Motives, R. A. Lundin and F. Hartman, Eds. Boston, MA: Kluwer Academic Publishers, 2000. [3] B. Egginton, "Multi-national consortium based projects: Improving the process," International Journal of Project Management, vol. 14, pp. 169-172, 1996. [4] G. Bachy and A.-P. Hameri, "What to be implemented at the early stage of a large-scale project," International Journal of Project Management, vol. 15, pp. 211-218 1997. [5] R. A. Sprague, K. J. Singh, and R. T. Wood, "Concurrent engineering in product development," IEEE Design & Test of Computers, 1991. [6] H. C. Zhang and D. Zhang, "Concurrent engineering: an overview from manufacturing engineering perspectives,“ Concurrent Engineering: Research and Applications vol. 3, pp. 221- 236, September 1995. [7] G. Probert, "Projects, people and practices," Engineering Management Journal pp. 141-6, June 1997. [8] R. P. Smith, "The historical roots of concurrent engineering fundamentals," IEEE Transaction on Engineering Management, vol. 44, Feb 1997. [9] H. Kerzner, Project management – a systems approach to planning scheduling and controlling, 8th ed. New York: John Wiley, 1998. [10] Wysocki R and R. McGary, Effective project management, 3rd ed. New York: John Wiley, 2003. [11] N. Hardie, "The prediction and control of project duration: A recursive model," International Journal of Project Management, vol. 19, pp. 401-409, 2001. [12] M. J. Schmidt, "Schedule monitoring of engineering projects," IEEE Transaction on Engineering Management, vol. 35, May 1988. [13] B. Elphick, "Linking the use of resources to project progress," Engineering Management Journal pp. 279-282, 1992. [14] A. S. Pillai, A. Joshi, and K. S. Rao, "Performance measurement of R&D projects in a multi-project, concurrent engineering environment," International Journal of Project Management, vol. 20, pp. 165-177, 2002. [15] C. Lohman, L. Fortuin, and M. Wouters, "Designing a performance measurement system: A case study,"European Journal of Operational Research vol. 156, pp.267-286, 2004. [16] M. Bourgault, E. Lefebvre, L. A. Lefebvre, R. Pellerin, and E. Elia, "Discussion of metrics for distributed project management: Preliminary findings," in Proceedings of the 35th Hawaii International Conference on System Sciences, Hawaii, 2002. [17] D. S. D. Jacob and D. Associates, "Forecasting project schedule completion with earned value metrics," The Measurable News 2003, vol. 1, pp. 7-9, March 2003. [18] F. S. F. Patrick, "Getting out from between Parkinson's rock and Murphy's hard place," PMNetwork, vol. 13, pp. 49-51, 1999.
Bibliography
8
