Project progress management under resource constraints and uncertain environmental factors

As a result, this method provides a more accurate project schedule and a simple and effective control mechanism for project progress management under resource constraints and uncertain e

Project Progress Management under Resource Constraints

and Uncertain Environmental Factors

（資源制約と不確実な環境因子を考慮したプロジェクト進捗管理

に関する研究）

by LUONG DUC LONG

A DISSERTATION

Submitted in partial fulfillment of requirement for the degree

DOCTOR OF PHILOSOPHY

(Information Science and Control Engineering)

Graduate School of Engineering

Nagaoka University of Technology

Japan- August 2007

Abstract

Project progress management is one of the most crucial aspects of project management Resource constraints such as limited numbers of workers or machines, and uncertain environmental factors such as weather conditions or management conditions make project progress management complicated and difficult In this research, a new method to effectively manage the progress of projects under resource constraints and uncertain environmental factors is developed The method is implemented in three logically interactive and iterative stages

In the first stage, the project is modeled by an activity network, which includes many activities that are connected by technological relationships, and often require many types of resources to be implemented To solve project-scheduling problem under resource constraints, the method takes into account the trade-off between the duration and resource consumption of activities The objective is to find the optimal duration of

each activity, so that project duration is minimized To find a satisfactory solution, a

genetic algorithm is introduced to determine the optimal activity durations, and priority heuristic rules are applied to find the start times of the activities The method provides

an initial schedule that often gives the minimum project duration and satisfies all resource constraints and technological relationships

In the second stage, using the initial schedule, the proposed method introduces a fuzzy activity network to model the project In this network, activity durations are estimated by fuzzy numbers By effectively using this fuzzy activity network, the proposed method calculates and analyzes the path durations and project duration under the influence of uncertain environmental factors Then, these estimated values are used

to develop a better schedule for project execution

In the third stage, the proposed method simply controls project progress using the fuzzy activity network and information about actual progress The method dynamically updates the schedule during project execution, so that the schedule reflects actual progress The method provides an early warning to the manager when there is some problem with project progress so that damage can be reduced

The three above stages of the proposed method are repeated until the project is completed.

As a result, this method provides a more accurate project schedule and a simple and effective control mechanism for project progress management under resource constraints and uncertain environmental factors This method is practical and easy to use and helps the manager to make the effective managerial decisions The proposed method has been shown to be flexible and powerful in practical progress management

The outline of the dissertation

This dissertation includes six chapters as following

Chapter 1- In this chapter, the introduction of project progress management under

resource constraints and uncertain environmental factors is given

Chapter 2- This chapter provides an overview of project progress management in

project management

Chapter 3- In this chapter, we newly develop a procedure named (P1) to flexibly and

powerfully solve project scheduling problem under resource constraints such as limited numbers of workers and machines

Chapter 4- In this chapter, a new procedure named (P2) to effectively calculate and

analyze the project schedule under effects of uncertain environmental factors such as conditions of weather, equipment, and management, is established

Chapter 5- This chapter describes development of a new method for project progress

management under resource constraints and uncertain environmental factors The method includes both procedure (P1) and procedure (P2) Chapter 6- This chapter summarizes concluding remarks, limitations of proposed

method and future researches

Contents

Page

Abstract……… i

Contents……… iii

List of figures……… viii

List of tables……… x

Chapter 1

Introduction

1.2 Project progress management under resource constraints

1.3 Practice of project progress management in Viet Nam 7

1.4 Problem description of project progress management under resource

constraints and uncertain environmental factors 10

1.4.1 Problem 1- The project scheduling problem under

1.4.2 Problem 2- Project scheduling problem under effects of

1.4.3 Problem 3- Problem of project controlling 10

1.5 Solution strategy for project progress management 11

1.5.3 General outline of the proposed method 13

Chapter 2

An Overview of Project Progress Management

2.1 Definitions of project, project management and project progress

2.1.4 Project progress management in the project life cycle 20

2.2 Literature review of project scheduling under resource constraints 21

2.2.1 Resource constrained-project scheduling with

2.2.2 Resource constrained project scheduling with

2.3 Literature review of project scheduling under uncertainty 26

2.3.1 Probability theory-based project scheduling 26

2.3.1.1 Probability theory based project scheduling

2.3.1.2 Probability theory based project scheduling

2.3.3 Fuzzy resource-constrained project scheduling 34

2.4 Literature review of project progress control 35

Chapter 3

Project Scheduling under Resource Constraints by Genetic

Algorithm

3.1.1 Project scheduling under resources constraints 38

3.1.2 Project scheduling under resources constraints with variable

activity durations (Problem 1) 39

3.2 Description of the resource-constrained project scheduling with variable

Project Schedule Management Considering Uncertain

Environmental Factors Using a Fuzzy Activity Network

4.2 Fuzzy activity network under the effects of uncertain environmental

factors 66

4.2.3 Fuzzy activity network under effects uncertain

4.3 The proposed procedure (P2) for project scheduling under uncertain

4.3.1 Component hierarchy in activity duration 70

4.3.3 Breakdown of uncertainty by factor conditions 73

4.3.6 Estimating project duration 82

4.3.8 Compatibility estimation between project duration

Chapter 5

Project Progress Management under Resource Constraints

and Uncertain Environmental Factors

5.1.1 Fuzzy set theory for modeling uncertainty 96

5.2 Detailed description of project progress management 98 5.2.1 Detailed description of resource constrained-project

scheduling problem with variable activity durations 98 5.2.2 Problem description of project progress management under

5.2.3 Problem description of project controlling 102 5.3 Project progress management under resource constraints

5.3.2 Dealing with resource constraints in project

5.3.3 Dealing with uncertainties in project scheduling

5.3.3.1 Strategy to deal with uncertainty 103 5.3.3.2 Project duration with high agreement index 105

5.3.3.4 Relationship between the project duration

and the agreement index 107

5.3.4 Controlling the project progress (Stage 3) 107

5.3.4.1 Revision of activity durations at the execution phase 107

5.3.4.1 Updating the planned schedule during execution

List of Figures

Fig 1.1 Project performance under resource constraints and

Fig 1.3 Framework of the new method for project

Fig.3.2 Gantt chart (a) and Resource graph (b) of an activity network

Fig.3.3 Schematic diagram of the optimization process for RCPSP

Fig.3.4 The relationship between activity duration and

Fig.3.5 Representation of assigned duration and the crossover

Fig.3.7 In the considered period (p), Act (5) is scheduled and Act(4) must be

Fig.3.8 In the next considered period (p+1), if the resource constraints

Fig.3.9 In the considered period (p+2), Act (4) is continuously considering to

Fig.4.5 The component hierarchy in an activity duration 70

Fig.4.9 The outline of the proposed procedure (P2) for project schedule

Fig.5.10 The relationship between the project duration and the

Fig.5.11 Transfer of the actual information into the project buffer

List of Tables

Table 3.2 The assigned durations, start times, and finish times

Table 3.5 The assigned durations, start times, and finish times

Table 3.7 Comparison of the proposed method to other

Table 4.2 The project duration of FNET, the proposed method,

Chapter 1

Introduction

1.2 Project progress management under resource constraints

1.3 Practice of project progress management in Viet Nam 7

1.4 Problem description of project progress management under resource

constraints and uncertain environmental factors 10

1.4.1 Problem 1- The project scheduling problem under

1.4.2 Problem 2- Project scheduling problem under effects of

1.4.3 Problem 3- Problem of project controlling 10

1.5 Solution strategy for project progress management 11

1.5.3 General outline of the proposed method 13

Chapter 1

Introduction

1.1 Introduction of project progress management

Managing projects dates back at least 4500 years (e.g., the builder of the pyramids

in Egypt are often cited as the world’s first project managers) [1]-[5] Nowadays projects, sets of activities which have a defined start point and a defined end state, and which pursue a defined goal and use a defined set of resources, come in many and various forms [4] The Apollo moon program, the construction of the Channel tunnel, the development of new products, and the construction of large office building are well-known examples of projects

The use of project management, which can be broadly defined as the process of managing, allocating and timing resources to achieve given objectives in an efficient and expedient manner, continuous to grow rapidly

The field of project management has taken tremendous strides forward in the past decades [1]-[7] In today’s competitive environment, it is crucial to deliver quality products on time and within budget A successful project is a project that it is finished

on time, within the budget, and according to the preset specifications However, cases where project deadline, budget, and/or specification are violated are quite common [4] Project management basically involves the planning, scheduling, and control of project activities to achieve performance, cost and time objectives for a given scope of work, while using resources efficiently and effectively [4] Project management has become very important for success of a project [1]-[7]

Project management includes many aspects such as quality management, progress (schedule/time) management, cost management, risk management, human management, contract/procurement management, communications management Meeting the three objectives (on time, within budget, according specification) while using resources efficiently lies at the heart of project management and is all too often overlooked

In many projects, project delays are more heavily penalized than budget overruns [4] For example, Hewlett- Packard studied a high-growth market and found that if a

new product innovation project was six moths late, it could cause a 33% loss in profits, while if the project overran its cost budget by 50% but was completed on time, the loss was only 3.5% (p.6 [4]) In other examples such as construction projects, a deadline project duration created is very critical to make the contract between owners and contractors If a project is delayed beyond its due date, a financial penalty is incurred by the contractor Furthermore, the contractor incurs substantial overhead (both in the home office and at the job site) each day that a project is going The overhead is estimated about 11.3 of gross revenues [6] Beside the financial problem when a project

is delayed beyond its due date, the contractor’s reputation is also reduced in the next bid for other projects

Under such circumstances, it is not surprise then to see that minimizing project duration ranks as the prime objective in project management [1]-[4] Hence, it is often

identified by project management experience that the failure in progress management is

one of the most common causes of project failure (p.6 [4])

This research also focuses on progress management, which involves a development

of an initial project schedule, and execution of this initial schedule to achieve project objectives

1.2 Project progress management under resource constraints and uncertainty

Fig 1.1 Project performance under resource constraints and

uncertain environmental factors

Resource Constraints

Uncertain environmental factor 2

Uncertain environmental

factor 1

Uncertain environmental factor 4

Uncertain environmental factor 3

PROJECT

Nowadays, new projects (e.g., construction projects) are dramatically increasing in developing countries These projects are often executed under resource constraints (e.g., limited numbers of skillful workers, and machines), and uncertain environmental factors (e.g., the weather condition, the general quality condition of labors, and the general quality condition of management), as simply shown in Fig 1.2

In today’s competitive environment, the demand to finish shorter project completion time is increasing This situation pushes project to permanently struggle to reduce project execution time The situation is aggravated by the combined effect of uncertainty and resource constraints

ƒ Resource constraints

In general, resources are classified into two main groups: renewable resources and nonrenewable resources

- Renewable resources such as workers, machines are resources whose

availabilities are limited quantities but can be renewable from period to period Only the total resource use at every time instant is constrained Typical examples of renewable resource include workers, machines, tools, equipments, space Consider the case of skilled labor: the number of skilled workers available to work on the project each day is limited, although no constraint is placed on the number of days skilled workers may be used

- Nonrenewable resources are resources whose total amounts are limited over the

project life This means that nonrenewable resources are available on a total project, with limited consumption availability for entire project Money is perhaps the best example of a nonrenewable resource: overall project costs are frequently limited to a fixed predetermined contract price Other examples include material and energy, because their total consumptions over the whole project duration are limited to the certain values

- Other resources can be considered as a combination of a renewable and a

nonrenewable resource These resources are constrained per period as well as for the overall project Capital with restricted project cash flow and limited total cash amount is a typical example

In project management, a schedule of workers/ machines represents a major

challenge because skilled workers (e.g., carpenters, and laborers in a construction

project) represent limited and expensive resources, which must be used effectively for the project to complete on time and within the budget While the proper use of material

is also important to development of an effective schedule, the effective deployment of workers (or machines) is the most critical Many studies have devoted for project scheduling under renewable resources (workers and machines) [8]-[31] Hence,

renewable resources are further investigated in this study

For project scheduling with renewable resources, resource availability assigned to a project is limited (called as resource constraints, such as limited numbers of labors or machines) and often not sufficient to concurrently do all eligible activities In this situation, the decision must be made about the start times of activities in order to

increase project completion time as little as possible From that point of view, we have

the problem of project scheduling under resources constraints (referred as constrained project-scheduling problem (RCPSP)), which arises when there are the

resource-limited amounts of available resources Scheduling objective is to extend project duration as little as possible beyond the original critical path duration in such a way that resource constraints are met In this process, both critical and non-critical activities

are shifted

A schedule without considering resource constraints often is not realistic This is because activities cannot be started if not enough resources for their requirements As a result, the project is done in an arbitrary schedule that often causes project delay, cost overrun, bad procurement, and dispute

Therefore, project scheduling under resource constraints is very important problem

in project management

ƒ Uncertain environmental factors

During execution of a project, activities are subject to considerable uncertainty that may lead to disruptions This uncertainty may stem from a number of possible sources: activities may take more or less time than originally estimated, weather conditions may cause severe delays, machines or equipments may become unavailable, material may arrive behind schedule etc Uncertainty in a project usually leads to schedule delays, cost overruns, and productivity losses, dispute between many parties

In other hand of considering resource constraints, it is very important to cope with problems of project schedule uncertainty for project progress management This is because a proper uncertainty analysis in project scheduling will provide a reliable

schedule in which risks of an unexpected schedule overrun are significantly reduced Project progress management under uncertainty has been intensively paid attention by many researches [43]-[105]

In general, projects normally are executed in an environment characterized by varying degrees of uncertainties Uncertain environmental factors normally affect project activity performance and cause uncertainty in duration estimations of activities When particular environmental factors (e.g., weather, labor skill, or management quality) have the same impact on multiple activities on the same path in the project

activity network, the activity durations tend to vary together in the same tendency, and

so the variations of these activities will be correlated [49][57][105] For example, two

activities on the same path are sensitive to the “weather” factor (an activity is sensitive

to a particular factor if its duration will increase or decrease when this factor condition changes), and the “weather” factor is considered to have three conditions (better-than-expected, expected, or worse-than-expected) When the weather condition is “better-than-expected”, both activity durations will tend to decrease, and vice versa

The variations of these durations will affect the path duration and/or the project duration Hence, both path durations and project duration are changed by effects of uncertain environmental factors [49],[57],[105] Therefore, uncertain environmental factors have considerable effects on project schedule uncertainty

This is an important reason of the uncertainty analysis under effects of these factors

in project progress management for providing the reliable project schedule in which the risk of an unexpected schedule overrun is significantly reduced

Furthermore, due to uniqueness of projects, historical data about activity durations are often not available Probability distributions for the activity durations are unknown

As a result, activity durations are estimated by human experts For example: “The activity duration is clearly more than 2 days and less then 5 days; about 3 days is usual” Under this circumstance, the uncertainty analysis in project progress management becomes complicated and difficult

ƒ Project controlling in execution phase

After having an initial schedule in project planning phase, this schedule needs to be actually performed in execution phase Hence, project controlling is also a very important for project progress management As earlier described, project controlling is

to measure, check, update, and operate the project progress in order to meet the initial

objectives Therefore, project controlling aims to implement the project with the expected time and budget

Unfortunately, during project execution, delays and unexpected interruptions (e.g., bad weather conditions, equipment breakdowns, and management interference) often happen during project execution phase, and they strongly affect activity durations and project duration The durations of interruptions may be not deterministic durations in reality, and this causes difficulties in managing and controlling project progress

ƒ Project progress management under resource constraints and uncertainty

From such circumstances, these projects are facing with many problems due to resource constraints (e.g., limited numbers of skillful workers, and machines), and uncertainty in which uncertain environmental factors are the main uncertainty sources (e.g., the weather condition, the general quality condition of labors, and the general quality condition of management)

As a result, project managers are often unable to create the effective schedule and control it under resource constraints and uncertainty Therefore, it is very necessary to develop an effective method, which is able to consider resource constraints and uncertainty in both planning phase and execution phase

1.3 Practice of project progress management in Viet Nam

Vietnam is a dynamic economy along with other Southeast Asian countries in the last few years Vietnam had an average growth in GDP of 7.1% per year from 2000 to

2004 The GDP growth was 8.4% in 2005, the second largest growth in Asia, trailing only China's Government figures of GDP growth in 2006, was 8.17% (Bloomberg News, August 10, 2006) November 2006, Vietnam became WTO's 150th member, after 11 years of preparation, including 8 years of negotiation This should provide an important boost to Vietnam's economy However, WTO accession also brings serious challenges, requiring Vietnam's economic sectors to open the door to foreign competition

For recent years in Viet Nam, construction projects with new technologies have been dramatically increased according to the economy growth, so the dependence of the project success on modern equipments and machines, special material, skillful workers,

competence of project management teams has became very important Hence, resource constraints (e.g., the limited numbers of skillful workers, modern equipment or machines) and uncertainty (e.g., inaccurate time estimating, inaccurate cost estimating) have strongly affected the project duration and the project success

* There were many projects in Viet Nam were delays due to many causes in which resource problems (such as shortage in material, money, labor supply, equipment availability) and uncertainty (unpredicted change) are the most critical ones The main damage of delays in projects was recognized as time overrun, cost overrun (normally when the projects are delayed, they are either extended or accelerated and therefore, incur additional cost), dispute of contractor and owner, litigation between different parties

Fig 1.2 Hai Van Tunnel Project, Viet Nam (Source http://www.hazama.co.jp/)

The Hai Van Tunnel, the longest tunnel in Southeast Asia at 6.28 km, lies on the highway between the two big cities of Da Nang and Hue in central Vietnam The Hai Van Pass has long been recognized as a major bottleneck The concept of a tunnel to replace the long route over the top had long been sought The main tunnel is 11.9 meters wide A second tunnel running alongside the main tunnel is currently used for maintenance and emergencies It is connected to the main tunnel by 15 cross-tunnels The tunnels have lighting, fire alarms, communication, water supply and treatment, ventilation fans, and radio broadcast systems The project's major constructor was a Hazama (Japan)-Cienco 6 (Viet Nam) joint venture The project was started as 27 August 2000 and the planned project completion date as January 2005 However, the

project was delayed around 6 months, and the actual project completion date was 5 June

2005 The project was delayed because of problems such as lack of considering uncertainty in the project schedule, insufficient resource allocation, and several labor strikes due to increasing working hours (overtime), lowering pay, and dangerous work conditions

The project of Rach Mieu Bridge in Tien Giang and Ben Tre provinces can be an adequate representative project The project is BOT project that is a form of private investment in public projects Ben Tre is one of the poorest provinces in the Cuu Long (Mekong) Delta, despite fertile land and the potential to develop fisheries One of the main factors hindering development in Ben Tre is poor transport To reach the province from Ho Chi Minh City, travelers must take the Rach Mieu Ferry When construction of the bridge is finished, it will take only a few minutes to cross the Tien Giang River, and the Rach Mieu bridge project will open a new development period of Ben Tre and Tien Giang provinces, realizing the long dream of residents on the two banks of the Tien River The project was started in May 2002, and the expected completion was in Dec

2005 However, up to now, the project is still performing even of the excess of the basic deadline The main reason is that improper planning due to the lack of resources, inadequate client’s finance and payment for completed work and many generated other problems Owner has been to extend the new project completion as Dec 2007 (two years delayed) with the increment of the estimated total project cost from around 37.9 million USD to around 44.1 million USD

* Damage of delays in the above projects (such as cost overrun, dispute of contractor and owner, litigation between different parties) was serious for both the contractor and the owner Delay analysis has pointed out that the bad schedule due to insufficiently considering resource constraints and uncertainty is one of the main causes Resource constraints and uncertainty are continuously predicted to be very important for the success of projects in the future

* In summary, main causes of delays in projects are a poor schedule and an improper control in project progress management Contractors often cannot make a practical and workable schedule at planning phase, and so cannot have the good schedule Hence, a poor scheduling at planning phase of a project often causes delays at execution phase, and then the improper correction will made the problem become more serious Of course, these problems are not only for the Vietnamese construction industry but also for those in other developing countries

1.4 Problem description of project progress management

1.4.1 Problem 1- Project scheduling problem under resource constraints

In almost previous research works for RCPSP, activity durations are often fixed initially and then the optimization process is performed to find the optimal start times

of these activities in order to meet the resource constraints [8]-[25], [27]-[30] However, in general, durations of activities are often variable in the ways which dependent on the trade off between durations of activities and their resource consumptions Thus, it is highly probable that these fixed durations lead to unexpected results (Long et al., 2006) [31]

ƒ Hence, the problem (1) - is to determine the optimal durations of

activities and their start times, so that project duration is minimized

under resource constraints

1.4.2 Problem 2- Project scheduling problem under influence of uncertain environmental factors

Many previous project scheduling methods [43]-[56], [83]-[86], [88]-[104] assumed that uncertainty in each activity is independent others In practice, the uncertain environmental factors such as weather condition factor, management condition factor, and general equipment condition factor often considerably affect many activities in a project

When a factor has the same impact on multiple activities, these activity durations tend to vary together in the same tendency, and so the variations of these activities will

be correlated These variations of activity durations normally affect the project

duration, and significantly cause the considerable uncertainty in the estimation of project duration

ƒ Therefore, the problem 2- is that how to calculate and analyze path durations and

project duration under uncertain environmental factors, to provide the more

reliable project schedule, so that risk of schedule overrun is significantly reduced

1.4.3 Problem 3- Problem of project controlling during execution phase

The initial schedule needs to be actually performed to achieve project During execution phase, the schedule often suffers delays and interruptions The manager needs

to measure differences between the actual schedule and the initial schedule, so that corrective managerial actions can be made

Previous fuzzy activity network methods [83]-[86], [88]-[104] are also not capable

to provide an appropriate tool to estimate lengths of delays/ interruptions and their effects on the remaining schedule under resource constraints As the result, many projects have been failed due to lack of considering resource constraints and uncertainty

in project execution

ƒ Therefore, the problem (3)- is that how to measure actual progress, and how to

estimate the uncertain lengths of delays and interruptions, and their effects on the remaining schedule under resource constraints, and how to provide the better schedule for the actual progress (we refer to this problem as the problem 3)

1.5 Solution strategy for project progress management

1.5.1 Fuzzy Theory for Modeling Uncertainty

Typically, in project progress management under uncertainty, there are two types of methods (some methods based on probability theory, or some methods based on fuzzy set theory)

Hence, to model uncertainties into activity durations in a project, project managers can select probability approach or fuzzy approach The probability approach is suitable for those situations where similar activities have been completed many times in the past and the information has been recorded for statistical analysis [43]-[73] The fuzzy approach is an alternative approach for project scheduling [80]-[105], and it may be suitable in following situations:

ƒ Project managers do not have enough information to obtain probability distributions of activity duration due to the non-repetitive or even unique projects, lack of repetition in activities, and different execution conditions at activity level between one project and another Managers have power to steer performance of activities, so they often make their estimations about activity duration based on subjective or incomplete information As a result, probability distributions for activity durations are unknown As activity durations are

estimated by experts, project management is often confronted with judgmental statements that are vague and imprecise [4] For example, the duration of an activity is clearly more than 2 days and less than 5 days, about 3 days is usual

ƒ Especially, in execution phase of project, probability theory cannot be used because many unexpected events may occur The manager will have to rely on his knowledge about actual situations to subjectively revise estimation of activity durations

ƒ Fuzzy set theory provides project managers a tool that is capable of dealing with uncertainties inherent in nature of delay analysis process and allows users to make assessments using linguistic terms, which suits actual practices commonly use in industry

* From the above points of view, we believe that fuzzy set theory is a more reliable alternative for schedule management in these situations

ƒ We newly develop the procedure (named as P1) in Chapter 3 Procedure (P1) is used to solve the project-scheduling problem under resource constraints (the above problem (1))

ƒ We establish the new procedure (named as P2) in Chapter 4 Procedure is (P2) used to deal with uncertainty in project scheduling under effects of uncertain environmental factors (the above problem (2))

ƒ From the use of Procedure (P1) and Procedure (P2), we newly develop the method to deal with both resource constraints and uncertainty in project progress management (all the problem (1), (2), and (3)) in Chapter 5

1.5.3 General outline of the proposed method

The outline of the proposed method for project progress management under resource constraints and uncertain environmental factors is shown in Fig 1.3, and will be described in Chapter 5 The proposed method includes two new procedures: the new procedure (P1) in Chapter 3 is used to solve the project-scheduling problem under resource constraints, and the new procedure (P2) in Chapter 4 is used to deal with uncertainty under effects of uncertain environmental factors

1.6 Scope of the dissertation

The scope of this dissertation can be summarized as following:

ƒ This study focuses on time management in project management Other aspects in project management such as cost management, quality management,

Dynamically update the project schedule,

if the buffer penetration is significant

Input data

Stage.2- Project scheduling under

uncertain environmental uncertain

factors (Procedure P2)- Chapter 4

Stage.1 – Project scheduling under

resource constraints (Procedure

P1)-Chapter 3

Stage.3- Controlling the project

progress by measuring buffer size

Project completion

Chapter 5-

The proposed

method

Fig 1.3 Framework of the proposed method for project progress management

communication management, procurement management are not considered in this research

ƒ This study only focuses on renewable resources such as workers or machines, which are the most important resources in project management

ƒ This study will be applied to manage progress of projects in which activity durations are estimated by experts, and project execution are affected by uncertain environmental factors (e.g., construction projects) Construction projects often do not have statistical data, and they are normally executed in environments affected by many uncertain factors For example, in bridge construction projects, activity durations are often estimated by analysts, and they are affected by uncertain environmental factors such as weather, materials delivery, labor, and equipment quality

1.7 Organization of the dissertation

This dissertation includes six chapters as following

ƒ Chapter 1- In this chapter, the introduction of project progress management under resource constraints and uncertain environmental factors is given This chapter also describes problems, solution strategy for project progress management, and scope of the dissertation

ƒ Chapter 2- This chapter provides an overview of project progress management

in project management, and an integrated survey of project progress management in literature review is also described

ƒ Chapter 3- In this chapter, we newly develop a procedure named (P1) to flexibly and powerfully solve project scheduling problem under resource constraints such as limited numbers of workers and machines The procedure (P1) is a new integration of a genetic algorithm and priority heuristic rules to find a satisfactory solution The result from (P1) is a desirable initial schedule that has minimum project duration, and satisfies all resource constraints

ƒ Chapter 4- In this chapter, a new procedure named (P2) to analyze the project schedule under effects of uncertain environmental factors such as conditions of weather, equipment, and management, is established The procedure (P2) effectively uses a fuzzy activity network, to calculate and analyze the path

durations and project duration Then, these estimated values are used to develop

a better schedule for project execution

ƒ Chapter 5- This chapter describes development of a new method for project progress management under resource constraints and uncertain environmental factors The method is implemented in three logically interactive and iterative stages In the first stage, the method solves the project-scheduling problem under resource constraints by (P1) In the second stage, the method analyzes the project schedule under effects of uncertain environmental factors by (P2) In the third stage, this method effectively controls the project progress by dynamically updating the schedule during project execution The three above stages are repeated until the project is completed It is shown that the proposed method is flexible and powerful in practical progress management

ƒ Chapter 6- This chapter summarizes concluding remarks, limitations of proposed method and future researches

2.1.4 Project progress management in the project life cycle 20

2.2 Literature review of project scheduling under resource constraints 21

2.2.1 Resource constrained-project scheduling with

2.2.2 Resource constrained project scheduling with

2.3 Literature review of project scheduling under uncertainty 26

2.3.1 Probability theory-based project scheduling 26

2.3.1.1 Probability theory based project scheduling

2.3.1.2 Probability theory based project scheduling

2.3.3 Fuzzy resource-constrained project scheduling 34

2.4 Literature review of project progress control 35

A project is defined as a temporary endeavor undertaken to create a unique product

or service that brings beneficial change or added value Temporary means that every project has a definite end Unique means that a product or service is different in some distinguishing way from all similar products or services [1] For example of unique, even many thousands of office buildings have been constructed, but each individual facility is unique- different owner, different design, different location and structure, different contractors, and so on

Another definition of project in [4] has widely accepted by users Project is a unique process, consisting of a set of co-ordinated and controlled activities with start and finish dates, undertaken to achieve an objective conforming to specific requirements including constraints of time, cost and resources (manpower, machinery, equipment, materials, energy, space, etc) [4]

Examples of projects include:

- Designing and constructing a building, a bridge, a tunnel, a highway

- Design, implement, testing a new computer system, new IT product

- Producing and directing a movie

- Producing a space shuttle

2.1.2 Definition of project management

A project is commonly considered as successful one when it is completed on time, within budget, and in accordance with specifications and to stakeholders’ satisfaction Project management is defined as the application of knowledge, skill, tools, and techniques to project activities in order to meet stakeholder’s needs and expectations from a project The project management knowledge areas include (PMBK,1996) [7]

1 Project integration Management describes the processes required to ensure that

the various elements of project are properly coordinated It consists of project plan development, project plan execution, and overall change control

2 Project Scope Management describes the processes required to ensure that the

project includes all the work required, and only the work required, to complete the project successfully It consists of initiation, scope planning, scope definition, scope verification, schedule development, and schedule control

3 Project progress management describes the processes required to ensure timely

completion of the project It consists of activity definition, activity sequencing, activity duration estimation, schedule development, and schedule control The project progress

management includes two integrated processes of project scheduling and project schedule controlling that mainly focus on the time aspect of project management

4 Project cost management describes the processes required to ensure that the

project is completed within the approved budget It consists of cost planning, cost estimating, cost budgeting, and cost control

5 Project Quality management describes the processes required to ensure that the

project will satisfy the needs for which it was undertaken It consists of quality planning, quality assurance, and quality control

6 Project human resource management describes the processes required to make

the most effective use of the people involved with the project It consists of organizational planning, staff acquisition, and team development

7 Project communication Management describes the processes required to timely

use project information It consists of communications planning, information distribution, performance reporting and administrative closure

8 Project Risk management- describes the processes concerned with identifying,

analyzing, and reporting to project risk It consists of risk identification, risk quantification, risk response development, and risk response control

9 Project Procurement Management describes the processes required to acquire

goods and services from outside the performing organization It consists of procurement planning, source selection, contract administration, and contract close out

* This research focuses on project progress management that is a crucial aspect in

project management The project progress management is a very difficult problem for both researchers and practitioners, so there were a vast number of studies for project progress management in the literature review [4]

2.1.3 Project progress management

Project progress management includes two interactive processes of project scheduling and project controlling that mainly focus on objectives about time Project progress management is also recognized as one of the most difficult challenges in project management by many researchers [1]-[7]

ƒ The first process of the project progress management is development of an initial project schedule or baseline schedule Project scheduling is to assign dates to activities (tasks) and to match the resources of equipment, materials and labor with activities (tasks) Project managers often use their project schedule as a basis tool to manage projects The initial schedule serves as a base schedule for planning the due day of project delivery, material procurement, communication and coordination between customer, owner, contractor, subcontractor, and supplier A good schedule can eliminate problems due to production bottlenecks, facilitate the timely procurement of necessary materials, and otherwise insure the completion of a project as soon as possible In contrast, a poor schedule often causes project delay, cost overrun, bad procurement, and dispute

ƒ The second process of project progress management is project-controlling by

using the initial schedule as a mean to achieve project objectives and required conditions Project controlling is to measure, check, update, and operate the project progress in order to meet the initial objectives

* According to complex characteristics of projects, importance of time objective in project is increasing in contracts As a result, an ability to early finish a project that is becoming a very important element in winning a bid, and the construction contract that

is often agreed with the heavily liquidated damages (e.g., heavy financial penalty) The project progress management will become very important for the success of projects Only a project, that is well scheduled and controlled under resource constraints and uncertainty, is finished on time successfully

2.1.4 Project progress management in the project life cycle

Generally, a project often passes through a life cycle that consists of different phases

as following [4]:

- The conceptual design phase: The need or requirement of project is identified

The options and alternatives are considered, and the feasibility of the best options is evaluated

- The next phase is the project definition phase In this phase, the project

objectives (three fundamental of project are time cost, and quality), project scope (what will and will not be done during the project), project strategy (what are going to be done and how to it to achieve the project objective) are defined The project is broken down into manageable work packages consisting of specific activities that are needed to perform in order to accomplish the project objectives This work breakdown process is called as Work Breakdown Structure that is a visual display of the project scope A summary view of the work breakdown process is given in Fig 1.1 Activity durations have to be estimated, and the resource requirement and availability, the precedence relationships among the

Project result

SUBDIVISION The project result is progressively divided until the components are identified in sufficient detail for effective planning and control

WOK PACKAGES - The components are further divided into work packages These work packages form the basis of cost control

ACTIVITIES – Work packages are represented in the network by one or more activities

Fig 2.1 Summary view of the work breakdown process [4]

activities have to be determined in sufficient detail Once the individual activities have been defined, they can be graphically portrayed in the form of a

project network

- The project enters planning and execution phases The project progress

management is mainly used in these phases The project progress management

includes two interactive processes of project scheduling and project controlling that mainly focus on the time aspect in project management Project scheduling aims to construct the feasible initial schedule that identifies the start and

completion times of the activities Project scheduling process also allocates resources to the different activities to optimize some objectives (e.g., to minimize the project duration) The initial schedule serves as a base schedule for planning the due day of project delivery, material procurement, communication and coordination between customer, owner, contractor, subcontractor, and

supplier Project controlling is very important in the execution phase This is

because when a baseline schedule has been established, the project schedule must then be turned into reality During the execution of the project, its progress must be monitored and corrective actions must be taken when necessary

- Finally, the project enters the termination phase that involves the delivery of

project results

2.2 Literature review of project scheduling under resource constraints

Project progress management has been widely recognized as a crucial aspect in project management Project scheduling is a very important process in project progress management, it is aimed at deciding when activities start (and finish), and how scarce resources are allocated to project activities, so that project objectives (e.g., project duration, project net present value) can be optimized

In project scheduling, activities are subjected to both technological precedence constraints such as sequential or parallel relationships between activates [1]-[5] For example of parallel relationship is that the electrical system can be installed by one crew

at the same time as the plumbing system is installed by a second crew

Furthermore, a project always needs many valuable and scarce resources (e.g., workers, and machines) to perform its activities for completing project These resources are often limited, so the activities must be scheduled under limited resources such as the

limited workers, limited equipments The schedule that did not consider the resource limitations is often delayed by the shortage of resources supply

Typically, resource constrained-project scheduling problem RCPSP arises when project activities are scheduled subject to both technological precedence constraints and resource constraints (the amounts of available resources are limited) [1]-[5] Hence, the scheduling objective is to minimizing the project duration (or to maximize the net present value of the project) in such a way that the precedence and resource constraints are satisfied The RCPSP is classified into two main groups as The RCPSP with the project duration objective, and the RCPSP with the project net present objective The RCPSP is recognized one of the most difficult challenges in project schedule management, also in operation research

Since the introduction of the Gantt chart, Critical Path Method (CPM) and Project Evaluation and Review Technique (PERT) which did not deal with resource-constrained project scheduling in their originality, the field of the RCPSP in project management has taken tremendous steps forward [1]~[6] Dealing with many variations of the RCPSP has been a challenge to both theoreticians and practitioners It has also led to a large number of researches [8]-[42]

2.2.1 Resource constrained-project scheduling with project duration objective

RCPSP with project duration objective can be solved by two approaches such as

exact analytical and heuristic methods In general, RCPSP is proved as NP-hard (NP denotes non-deterministic polynomial time), it means there are no known algorithms for finding optimal solutions in polynomial time Various exact analytical approaches have been formulated to solve the problem optimally such as integer linear programming, dynamic programming, and branch and bound attempt to find the optimum solution of the minimum project duration [1]-[4], [13]-[14], [23], [26], [30] However, exact mathematical methods are not effective for practical project because they have to face with the problem called combinatorial explosion due to large project

in real life (Kolisch et al, 1999) [21]

As a result, most researches have been devoted to efficient heuristics for finding near optimal solution for the RCPSP in practice Although heuristic approaches may not find the optimal solution for RCPSP, they provide reasonable solutions in practical time (Kolisch 1996, 1999) [16], [21] Heuristic rules are simple to understand, and very inexpensive to be used in computer programs Typically two standard heuristic methods

for the RCPSP have been used in literature such as the serial methods (Kelly,1963) [8] and the parallel methods (Wiest, 1964) [9]

ƒ In the serial heuristic method, all activities are sorted by network sequence and

by certain other priorities, and then each activity is scheduled one at a time (serially) in an order of a sorted list When an activity cannot be scheduled at its early start time (depending on its precedent activities) due to lack of resources, the activity start time is delayed until it meets its resources requirement

ƒ On the other hand, the parallel method schedules several activities at the same time (in parallel) At every given time interval, eligible activities are searched and sorted by certain priorities, and then they are scheduled in the order of the sorted list When an activity cannot be scheduled due to lack of resources, the activity is reconsidered in the next time interval

There are various serial and parallel heuristic approaches have been suggested, which generally employ intuitively logical rules of thumb designed for solving RCPSP in practice [8]-[12],[15]-[25], [27]-[28], [30] For example, priority rules based on network features, resource relationships, and activity criticality for the high priority activities to schedule under resource constraints

Meta heuristic methods such as Genetic algorithm (GA), Ant Colony algorithm, Tabu search method have also developed for solving the RCPSP in practice as followings:

Genetic algorithm (GA) is stochastic search techniques based upon the mechanism of natural selection and natural genetics GA is recognized as the most effective technique for determining optimal solutions especially when the problem domain is large and complex [15], [17], [21], [25], [28], [32] Genetic algorithm (GA) is very effective for RCPSP for large projects, and there are many efforts to apply GA for optimizing the scheduling problem [15], [17], [21], [25], [28]

Paul R.T et al (1998) [19] presented a Tabu search approach for the RCPSP in which

a forbidden list was created to reduce the solution space in his method

Ant Colony method (Merkle, 1999) [24] is an artificial intelligent algorithm based

on ants that find a good solution by putting some amount of pheromone on edges of path The following ants of next generations are attracted by the pheromone so that they search in the solution space near previous good solutions In addition to the pheromone

values, the ants will usually be guided by some problem-specific heuristic for evaluating possible decisions

2.2.2 Resource constrained- project scheduling with project net present

objective

In some cases, another objective of contractors and clients is to reduce costs and to maximize their profit represented by the net present value (NPV) NPV is the most frequent criterion used in project scheduling where financial aspects of project are taken into account This criterion is calculated with cash flows generated by project

activities A cash flow can be positive (client’s payments) or negative (the expenses caused by workers, equipment, raw materials) NPV of the project is considered when time value of money is taken into account Project scheduling with maximal NPV has been emphasized by several researchers and practitioners [34]-[42] In literature review, several methods have been proposed for project scheduling with maximal NPV These methods can be classified into two main groups (the first group for unconstrained resource project scheduling, and the second group for constrained resource project scheduling with NPV)

* The first group is very important for development of the second group, so there are many research works that have been developed for unconstrained resource project scheduling with NPV The pioneering work of Russell A H (1970) [34] was the first to consider how to maximize the net present value (NPV) of cash flows in a network Russell considered the unconstrained problem where both positive and negative payments occur as events in the project are competed He transformed the nonlinear objective function into a linear one by approximation using the first term of the associated Taylor series expansion The dual of the reformulated a model is a minimum cash flow problem

Grinold (1972) [35] added a project deadline to the model and used a transformation

of variables to show the project scheduling problem with NPV objective The model is a linear program with the structure of a weighted distribution problem

Doerch and Patterson (1977) [36] presented a binary integer-programming model of the project-scheduling problem to maximize its present value

Smithh Daniels (1987) [38] also proposed the mathematic programming model to state the trade-off between material cost, capital constraints, and project net present

value The model has too many constraints and variables, and it is difficult to solve this model even with the small size projects

Elmaghraby et al (1990) [39] used Russell’s model to develop an algorithm that gives the optimal schedule for project scheduling with NPV objective However, this above method is not used to solve large projects in real life due to its complicated procedure

* The second group is the development of the first group They are heuristic methods that have been developed for solving constrained resource project scheduling with NPV [37], [40]-[42] They can be represented by major researches as following: R.A Russell (1986) [37] provided one of the first comparisons of heuristics for scheduling projects with resource constraints where the objective is to maximize project net present value Russell has introduced priority rules for selecting activities for resource assignment based upon information derived from the optimal solution of the unconstrained problem He has evaluated five heuristic rules (three rules based on information of the optimal solution to the unconstrained problem in Russell A.H., and two heuristic rules are Minimum slack rule and Minimum latest finish time rule)

Padman R et al (1997) [40] presented the heuristic method for RCPSP with NPV, their proposed method has used a heuristic procedure with embedded priority rules that uses information (such as dual prices) from the revised solution of a relax optimization model (This relax model is the unconstrained NPV project scheduling problem in [34]) The priority rules that were used are maximum tardiness penalty rule, opportunity cost

of scheduling rule, opportunity cost of resource rule, net opportunity cost of scheduling rule

Baroum and Patterson (1996) [41] developed the heuristic procedures that assign weights to activities based upon cumulative search locally for obtaining improved solutions of the maximization of net present value in a network problem The heuristic rule is inherent simplicity and easy to apply in the practice The heuristic rules often produce improved results over other approaches when maximum net present value solutions are desired

Dan Z et al (1999) [42] applied the Tabu search metaheuristic procedure for the RCPSP with cash flows Strategies for neighborhood generation and candidate selection that exploit the special features of the problem are combined with a simple heuristic start procedure

ƒ However, in almost of these previous research works about RCPSP with project duration objective / with project present value objective, the trade-offs between activity durations and resource consumptions are not considered In the practical, the durations of activities are often variable in the certain ranges which dependent on their daily resource consumptions

ƒ Hence, we need to determine the optimal activity durations and start times of these activities so that the project duration is minimized (or in other case, so that the net present value of project is maximized) The problem is very hard to solve due to the intractable nature of this problem Consequently, one is motivated to develop well-performing heuristic procedures This problem is a motivation for the proposed method in Chapter 4

2.3 Literature review of project scheduling under uncertainty

In literature review of project scheduling under uncertainty, there are two main groups which can be used to model uncertainty in project activity durations: probability theory-based methods and fuzzy set theory- based methods

2.3.1 Probability theory-based project scheduling

The probability approach is suitable for projects where similar activities have been completed many times in the past, and the information has been recorded for statistical analysis There are many probability theory –based scheduling methods for analyzing project schedule uncertainty These methods can also be classified into two main groups such as scheduling methods with / without using buffers, where a buffer is a reserved amount of time that is inserted in the project schedule to protect this schedule from potential disturbances

2.3.1.1 Probability theory- based project scheduling without using buffers

Project Evaluation and Review Technique (PERT) is a network-based scheduling method It was developed during the design and construction of Polaris Submarine in the USA in the 1950 (Malcom, 1959) [43] PERT has been recognized as the probability-based scheduling method used most widely for project management PERT

is based on probability theory to estimate activity durations by using a three-point estimate (pessimist, most likely, optimist) In PERT, the central limit theorem is applied

as follows: The distribution of the sum of the expected durations of the activities along the critical path is approximately normal, particularly as the number of activities increases The expected duration of each sum is equal to the sum of the expected durations Similarly, the variance of each sum is the sum of the variances However, in

PERT did not give the correct project duration when there are merge-event bias problems The merge-event bias problem is occurred in a network where there is no a

single dominant path and several competing paths (with overlapping probability density functions) may determine the project duration

Van Slyke (1963) [44] proposed the use of Monte-Carlo simulation that can tackle the merge-event bias problem in the project network

In the pioneering work, Eisner (1962) [45] proposed a generalized network approach

to the planning and scheduling of a research project In his method, the failure or success probability of an event is considered

Elmaghraby (1964) and (1970) [46], [47] also proposed the algebra method for generalized activity network GAN which considered branch and loop in a network In GAN, each arc (activity) is defined by vectors of three parameters such as: a probability that arc will be realized, duration with a probability density function, a cost function that may or may not depend on the duration If all arcs are realized with certainty, the GAN model reduces to the PERT model If further more, all durations are fixed constraints, the model further reduces to the well-known critical path method (CPM)

Pritsker A.A.B (1966) [48] proposed the stochastic network method as Graphical Evaluation and Review Technique (GERT) in which a node in a stochastic network consists of an input (receiving, contributive) side and an output (emitting, distributive) side Three logical relations on the input side and two types of relations on the output side will be considered The three logical relations on the input side are “exclusive-or logic”, “inclusive-or logic”, and “and logic” The output side, the two relations are defined as deterministic and probabilistic logic output Hence, GERT can additionally model uncertainty in the precedence of activities GERT is the stochastic project network that has a stochastic evolution structure and feedback GERT did not calculate criticality in activities, and did not consider the resource constraints in its originality Carr R (1979) [49] proposed the simulation method for considering correlation between activities in network, in order to estimate project duration

Dawson et al (1994) and (1995) [50], [51] revised the traditional GAN with activity

on the arc to become the generalized activity-on-the node networks for managing uncertainty in projects His model is more flexible to deal with the complex project Carlos and Hadipriono (1993) [52] investigated the probabilistic network technique method (PNET) with PERT The basis of this approach is the following two inequalities that provide a lower and upper bound for the true cumulative probability of project completion ( P[T t] [P T t] [P T i t]

to be critical and sufficiently independent to be included in the calculation of the lower bound Further detailed information can be found in [52], [53]

Martinez (1996) [53] proposed simulation system (named as State and Resource Based Simulation of Construction Processes-STROBOSCOPE) that allows dependence and correlation between activity durations as well as control over precedence and the selection of paths Activity durations and sequencing can be defined in terms of the dynamic information that becomes available as a project evolves and includes the actual start date and duration of activities already started STROBOSCOPE can solve the merge-event bias problem in PERT Furthermore, STROBOSCOPE is powerful for the simulation of common processes to construction engineering such as project schedule simulation, operational simulation and queue simulations

There have also existed many revisions of PERT method as following For construction project scheduling, Wayne D (1999) [54] introduced the simplified PERT method which reduces the level of effort required by conventional PERT because only two time estimates, rather than three, are required for each activity Simplified PERT produces shorter project durations, but greater project duration variances, than does conventional PERT The combination of these two effects results in similar project duration probabilities Deng et al (2003) [56] presented the new method for calculating path float in PERT Pierpaolo (2000) [55] used PERT-path technique for estimating the project duration in stochastic networks

Wang (2000) [57], [58] has proposed the model for evaluating NETworks under CORrelated uncertainty- NETCOR in which is a probability based- method uses the simulation process to capture the factor effects NETCOR, which is run by using STROBOSCOPE (Martinez, 1996), has incorporated the effect of correlation in network schedules and provides factors sensitivity information to support schedule risk management NETCOR has provided much insight information about project schedule uncertainty for project manager

Vahid Khodakarami (2005) [60] presented the Bayesian Network Model for project planning that addresses both uncertainty and causality in project management Bayesian network is used for situations that require statistical inference [61] The project manager

knows some evidences (or events) actually been observed, and wishes to infer the

probabilities of other data, which have not yet been observed By applying Bayesian rule in each affected node, he/she can influence other BN nodes via propagation, modifying the probability distributions

ƒ In summary, the main advance of these above methods is that they can analyze

the project schedule uncertainty However, in some cases, they cannot provide

the correct results (e.g., PERT in the merge-event bias problem), and they cannot

be applied in projects in which there is insufficient statistical data for using probability theory

2.3.1.2 Probability theory- based project scheduling using buffers

Goldratt (1997) [62] introduced the critical chain project management (CCPM) method which is based on the probability theory CCPM uses a deterministic schedule integrated by a buffer mechanism to deal with both resource constraints and uncertain activity durations CCPM removes the hidden safety in activity durations to protect from starting late on activities (Student Syndrome) and keeping busy for the entire activity duration (Parkinson’s Law) Then these safety times are placed at key points in the deterministic schedule as buffers to absorb uncertainties in project execution phase All activities are scheduled as late as possible so that costs are not incurred earlier than necessary The critical chain is the longest chain of activities that considers both activity and resource dependencies The project buffer is placed at the end of the critical chain to protect against exceeding the project deadline, and feeding buffers are placed at the intersections between any non-critical chains and the critical chain to protect it against