An Estimating Method for IT Project Expected Duration Oriented to GERT

Similar documents
E-learning Strategies to Support Databases Courses: a Case Study

Maths Games Resource Kit - Sample Teaching Problem Solving

A Study of Metacognitive Awareness of Non-English Majors in L2 Listening

Reinforcement Learning by Comparing Immediate Reward

Circuit Simulators: A Revolutionary E-Learning Platform

M55205-Mastering Microsoft Project 2016

Application of Multimedia Technology in Vocabulary Learning for Engineering Students

SAM - Sensors, Actuators and Microcontrollers in Mobile Robots

GACE Computer Science Assessment Test at a Glance

FEIRONG YUAN, PH.D. Updated: April 15, 2016

From Empire to Twenty-First Century Britain: Economic and Political Development of Great Britain in the 19th and 20th Centuries 5HD391

Shockwheat. Statistics 1, Activity 1

Unvoiced Landmark Detection for Segment-based Mandarin Continuous Speech Recognition

Multidisciplinary Engineering Systems 2 nd and 3rd Year College-Wide Courses

A student diagnosing and evaluation system for laboratory-based academic exercises

Xinyu Tang. Education. Research Interests. Honors and Awards. Professional Experience

A GENERIC SPLIT PROCESS MODEL FOR ASSET MANAGEMENT DECISION-MAKING

What s in a Step? Toward General, Abstract Representations of Tutoring System Log Data

Simulation of Multi-stage Flash (MSF) Desalination Process

Radius STEM Readiness TM

Improving Fairness in Memory Scheduling

Education: Integrating Parallel and Distributed Computing in Computer Science Curricula

Learning Methods for Fuzzy Systems

By Laurence Capron and Will Mitchell, Boston, MA: Harvard Business Review Press, 2012.

Malicious User Suppression for Cooperative Spectrum Sensing in Cognitive Radio Networks using Dixon s Outlier Detection Method

ZHANG Xiaojun, XIONG Xiaoliang School of Finance and Business English, Wuhan Yangtze Business University, P.R.China,

A Coding System for Dynamic Topic Analysis: A Computer-Mediated Discourse Analysis Technique

Pedagogical Content Knowledge for Teaching Primary Mathematics: A Case Study of Two Teachers

OPTIMIZATINON OF TRAINING SETS FOR HEBBIAN-LEARNING- BASED CLASSIFIERS

Document number: 2013/ Programs Committee 6/2014 (July) Agenda Item 42.0 Bachelor of Engineering with Honours in Software Engineering

Emergency Management Games and Test Case Utility:

COMPUTER INTERFACES FOR TEACHING THE NINTENDO GENERATION

Bayllocator: A proactive system to predict server utilization and dynamically allocate memory resources using Bayesian networks and ballooning

Computer Science. Embedded systems today. Microcontroller MCR

AC : DEVELOPMENT OF AN INTRODUCTION TO INFRAS- TRUCTURE COURSE

Specification of the Verity Learning Companion and Self-Assessment Tool

Courses below are sorted by the column Field of study for your better orientation. The list is subject to change.

Evaluating Collaboration and Core Competence in a Virtual Enterprise

Lab Reports for Biology

Evaluating Progress NGA Center for Best Practices STEM Summit

Chapter 4 - Fractions

Web-based Learning Systems From HTML To MOODLE A Case Study

Transfer Learning Action Models by Measuring the Similarity of Different Domains

The UNF Digital Commons

Course specification

Axiom 2013 Team Description Paper

To link to this article: PLEASE SCROLL DOWN FOR ARTICLE

Probability and Game Theory Course Syllabus

New Paths to Learning with Chromebooks

SITUATING AN ENVIRONMENT TO PROMOTE DESIGN CREATIVITY BY EXPANDING STRUCTURE HOLES

An Introduction to Simio for Beginners

MMOG Subscription Business Models: Table of Contents

University of Groningen. Systemen, planning, netwerken Bosman, Aart

Reduce the Failure Rate of the Screwing Process with Six Sigma Approach

Constructing a support system for self-learning playing the piano at the beginning stage

- INFORMATION TECHNOLOGIES AND TELECOMMUNICATION

Entrepreneurial Discovery and the Demmert/Klein Experiment: Additional Evidence from Germany

Process improvement, The Agile Way! By Ben Linders Published in Methods and Tools, winter

On the Combined Behavior of Autonomous Resource Management Agents

Operational Knowledge Management: a way to manage competence

Planning for Preassessment. Kathy Paul Johnston CSD Johnston, Iowa

Power Systems Engineering

The Method of Immersion the Problem of Comparing Technical Objects in an Expert Shell in the Class of Artificial Intelligence Algorithms

"On-board training tools for long term missions" Experiment Overview. 1. Abstract:

Cal s Dinner Card Deals

Australian Journal of Basic and Applied Sciences

Degree Qualification Profiles Intellectual Skills

AP Statistics Summer Assignment 17-18

Execution Plan for Software Engineering Education in Taiwan

ECE-492 SENIOR ADVANCED DESIGN PROJECT

E-Learning Based Teaching Material for Calculus in Engineer Training

Integration of ICT in Teaching and Learning

MIAO WANG. Articles in Refereed Journals and Book Volumes. Department of Economics Marquette University 606 N. 13 th Street Milwaukee, WI 53233

Multiple Measures Assessment Project - FAQs

BSc (Hons) Banking Practice and Management (Full-time programmes of study)

Using Proportions to Solve Percentage Problems I

Prairie View A&M University Houston, TX P.O. Box 519; MS 2220; Hilliard Hall (281)

Java Programming. Specialized Certificate

Economics at UCD. Professor Karl Whelan Presentation at Open Evening January 17, 2017

The Use of Statistical, Computational and Modelling Tools in Higher Learning Institutions: A Case Study of the University of Dodoma

Meet the Experts Fall Freebie November 5, 2015

Introduction to Simulation

The Round Earth Project. Collaborative VR for Elementary School Kids

E-Learning project in GIS education

Development of an IT Curriculum. Dr. Jochen Koubek Humboldt-Universität zu Berlin Technische Universität Berlin 2008

Business Administration

Completed applications due via online submission at by 11:59pm or to the SEC Information Desk by 7:59pm.

Chapter 10 APPLYING TOPIC MODELING TO FORENSIC DATA. 1. Introduction. Alta de Waal, Jacobus Venter and Etienne Barnard

Evaluation of Hybrid Online Instruction in Sport Management

ICT Strategy of Universities

Study Group Handbook

BUILDING CAPACITY FOR COLLEGE AND CAREER READINESS: LESSONS LEARNED FROM NAEP ITEM ANALYSES. Council of the Great City Schools

CONCEPT MAPS AS A DEVICE FOR LEARNING DATABASE CONCEPTS

Managerial Decision Making

Language Independent Passage Retrieval for Question Answering

Blended Learning Models and Lessons from the Field. Julia Freeland Fisher

Yizao Liu

Characteristics of Functions

Measurement & Analysis in the Real World

Analysis of Enzyme Kinetic Data

Transcription:

An Estimating Method for IT Project Expected Duration Oriented to GERT Li Yu and Meiyun Zuo School of Information, Renmin University of China, Beijing 100872, P.R. China buaayuli@mc.e(iuxn zuomeiyun@263.nct Abstract. Reworking in developing software exists widely in IT project management. But the current estimating methods on project duration cannot deal well with this problem. In this paper, some key concepts, such as adjusted activity, adjusted GERT network, expected project duration, expected critical path, are firstly defined. Then authors study estimating method on project duration for different rework modes, including serial, paratactic, and combined structure by encapsulating of activities. Finally, a general estimating method on IT project duration oriented to GERT is presented, and detailed steps are illustrated with an example project on IT software development. Keywords: Project Management, Project Schedule Estimation, Expected Duration L INTRODUCTION In project management, project schedule often faces with the problem how to know project duration under the restriction of resource. PERT (Program Evaluation and Review Technique) and CPM (Critical Path Method) are the most important project management techniques [1], which have been created out in the need of Western industrial and military establishments to plan, schedule and control complex projects [2]. GERT ( Graphical Evaluation and Review Technique ) is another important analysis technique, being widely using in engineering management and enterprise producing for strong ability to describe reworking in project management [1]. Reworking exists in IT project management widely, and GERT can describe it well. But current method to estimate project duration only can deal with determinate project network without reworking. So, the problem how to estimate project duration with reworking became key technique for the using of GERT. Currently, few such researches are made in the world. Very few research is made on estimating project duration oriented to GERT by simulating method [3]. But it is difficult to compute accurately project duration, especially when times of simulating is few, the error is more obvious. The paper tries to resolve the problem by encapsulating of activities. In next section, key concepts, such as expected duration and expected critical path etc, are presented. In section 3, author research on estimating method on IT project duration under all kinds of reworking cases, including serial, paratactic, and combined Please use the following format when citing this chapter: Yu, L., Zuo, M., 2007, in IFIP International Federation for Information Processing, Volume 255, Research and Practical Issues of Enterprise Information Systems II Volume 2, eds. L. Xu, Tjoa A., Chaudhry S. (Boston: Springer), pp. 1557-1566.

1558 Li Yu and Meiyim Zuo structure. General resolving method is given in section 4. In section 5, estimating method is used in a project example about the development of information system project. 2. EXPECTED CRITICAL PATH AND EXPECTED PROJECT DURATION Under the condition with reworking, critical path can not be determined because the number of reworking for certain activity is not determinate. So there may be a lot of critical paths which have same activities in a project network diagram with reworking. Since the number of activities consisting of a critical path is determinate and each activity has determinate duration, project network diagram is determinate. Project duration could be infinite when reworking of certain activity always happened. If no reworking happened, GERT network became determinate network and project duration is determinate. Project duration can be computed by CPM and PERT methods. But both of the methods are not adapt to GERT network with reworking. Although critical path and duration of the project based on GERT are not determinate, expect duration of each activity could be computed according to the probability distribution of the reworking. Then, according to expected duration of each activity, critical path of project can be computed by using CPM/PERT methods. Because this kind of critical path is based on expected duration of each activity, we call it as expected critical path of project, and corresponding duration as expected project duration. They are very useftil in project schedule management when reworking could happen. 3. ESTIMATING OF EXPECTED PROJECT DURATION FOE DIFFERENT REWORKING MODES 3.1 Including One Reworking Activity Here, being supposing that a project only includes one reworking activity A, and its duration is a, the probability of reworking is p. As shown in Fig.l. Now, how to compute project duration? Noting the expected project duration as D, so D = a + (p*d) (1) And 1-/7

An Estimating Method for IT Project Expected Duration Oriented to GERT 1559 START H ) ( J-^ ^ END DA=a Figure 1. Project Including One Reworking Activity 'J3_> ND Figure 2. Serial Reworking In the following, two extremities be analyzed. In this case, it means that no reworking exists, and expected project duration can be computed as following, (2) p=l In this case, it means that reworking always exists. As be showed in Equation (4), expected project duration is +00, 0 = ^ \,=+00 (4) Generally, bigger p shows higher reworking probability, so expected project duration is longer and D=D(p) increased with increasing of reworking probability p. Sometimes, it is usual that project allows top times of reworking. That is to say, after top number of reworking, the activity must be finished. The mode exists widely in project management practice. In this mode, computing expected project duration has practical meaning. Let's noting Dn as project duration while allowing n times of reworking for an activity, then the number list {Dn} (n=l,2...) has following equations, D 0= a D ] = a+pa D 2= a^pct^p a... Dr^=a+pa^...+p''a So, D \-p Especially, while ^ ^ +^, it means there is no limit for times of reworking. So

1560 Li Yu and Meiyun Zuo "^^,^->fco \ p l p This conclusion is consistent with above the resuh. 3.2 Including Lots of Reworking Activities 3.2.1 Serial Reworking As shown in Fig.2, a project includes two serial reworking activities A and B. Then And DA and DB can be computed as Equation (1) and Equation (2 ), So D,=b + (p,*d,) ^ D,=-^ i-p4 a b D = D^+Dg=- + - (6) I-/J2 I-/' 3.2.2 Paratactic Reworking As shown in Fig.3, a project includes two paratactic reworking activities A and B. Then D = msk{d^,d,} And DA and DB can be computed as Equation (1) and Equation (2 ), so I-P2 D,=b + ip,^d,) ^ D,=- Then D = m?oip^,d^} = m3x{-^, } (7)

An Estimating Method for IT Project Expected Duration Oriented to GERT 1561 Figure 3. Paratactic Reworking DB=b Figure 4. Combined Reworking 3.2.3 Combined Reworking As shown in Fig.4, project includes serial reworking and paratactic reworking. Among of them, activities A and B are paratactic, being serial with activity C. Then, expected project duration is D = max{d^,dj-\-dc According to Equation (6) and Equation (7), D be computed as following D=m^xp^,DJ+Dc =maxf-^, }+ ^ (8) i-p^ l-p^ 1--Pe 3.3 Reworking on Lots of Activities As shown in Fig.5, if activity B needs reworking, activity A also must rework. Reworking relates to two activities. According to Equation (1), then So D = - (9) ^-Pi

1562 Li Yu and Meiyun Zuo > END Figure 5. Reworking on Lots of Activities B KTART -(y 0--*-END Dc=c Figure 6. Multi-choices of Reworking 3.4 Multi-choices of Reworking In IT project management, there are often lots of path choices for one activity. As shown in Fig.6, after activity A is finished, reworking could be happened at once or activity B or C could be executed. That is to say, after activity A be finished, there are three path choices. In this case how to compute project duration? In fact, this case can be transformed to two reworking choices mode. Activity B and C can be encapsulated to one virtual activity D, as be shown in Fig.7. After encapsulating activity B and C, according to the probability theory, the duration of virtual activity D can be computed as following P2*b + p,*c Do = P2+P3 (10) So, project duration is D=D.+Il-- a p^ *b+pj *c a+p^ *b+p *c (11) l-z'i Pi+Pi Pi+Pi P2+P3 T=Dn <y* END Figure 7. After Encapsulating Activity B and C

An Estimating Method for IT Project Expected Duration Oriented to GERT 1563 Modified activity A A T=D^ Figure 8. Encapsulating Activity 4. ESTIMATING STEPS OF EXPECTED PROJECT DURATION ORIENTED GERT After discussing the computing method on expected project duration under different modes, we provide the general estimating steps of expected project duration oriented to GERT in the following. 4,1 Encapsulating Reworking Activities The key step in estimating of expected project duration oriented to GERT is to transform GERT network with reworking activities to traditional determinate network without reworking activities. So the first step is to modify GERT network by encapsulating reworking activities, as shown in Figure 8. Activity A with reworking is transformed to modified activity A* without reworking by encapsulating. We call activity A* as adjusted activity of the activity A. 4.2 Computing of Duration of Adjusted Activity ^ According to Equation (2), duration of adjusted activity A* is D : = - - ^ (12) We call it as adjusted duration of activity A. 4.3 Generating Adjusted GERT Network Doing as above two steps for all activities with reworking, we can generate new GERT network without reworking. We call the network as adjusted network diagram. 4.4 Computing of Expected Critical Path and Expected Duration After eliminating reworking activity by encapsulating and generating adjusted GERT network, expected critical path and expected project duration can be computed by traditional CPM/PERT methods.

1564 Li Yu and Meiyun Zuo 5. EXAMPLE For better explaining the above method, here we give a IT development project consisting of system design, buying hardware, software analysis, buying software, hardware examining, hardware setup, etc, as shown in Fig.9. Among these activities, two activities, hardware examining and system examining, could exist reworking. If there exists questions in hardware examining activity, hardware modifying activity happened. System examining happens according to trouble degree. All activities duration and reworking probability are shown in Fig.9. Now, we try to compute expected critical path and project duration. Figure 9. GERT Network of IT Project Hudwaie I > M Figure 10. Modified Activity M after Encapsulating 5.1 Encapsulating Reworking Activities In this example, there exist reworking in two activities. They can be encapsulated to two adjusted activities M and N, as shown in Fig. 10 and Fig.l 1. 5.2 Computing of Expected Duration of Modified Activity (1) Expected Duration of Adjusted Activity M According to Equation (2), expected duration of adjusted activity M can be computed as following,

An Estimating Method for IT Project Expected Duration Oriented to GERT 1565 D,+Dj, ^3 + 4 DM=- ^ 0.8 = 8.75 (2 ) Expected Duration of Modified Activity N Because activity K exist three path choices, according to Equation (10), expected duration of adjusted activity N can be computed as following, P3*9 + P5*18 4.5 ^, Di6,2i= = TTT = 7.5 0.6 0.6 Di3,2i= Di3,i4+ Di4,i5+ Di5,i6+D,6,2i =4+0+6+11.25=21.25 C::tKi. m^^^ I r Jr_ * y^"^ I 1.. I Figure 11. Modified Activity N after Encapsulating Hardware M i&75 Hanurare System Buyi^jfcABi 4 (Start \^/Y) Pg"g" -/^^ 0 SofH are Nete G= 3 So Aw are Hardnv Fi 6 w^- (0) ^- ^ 17.5 Figure 12. Modified GERT Network Without Reworking 5.3 Generating Adjusted GERT Network After encapsulating all reworking activities and computing their adjusted duration, we can get adjusted GERT network without reworking, as shown in Fig. 12. 5.4 Computing of Expected Critical Path and Expected Duration Using CPM techniques, we can get expected project critical path, A B C M~F N Then, its expected duration can be computed.

1566 Li Yu and Meiyiin Zuo D=30+4+3+8.75+6+21.25=73 (days) 6. CONCLUSIONS Reworking exists widely in IT project management. The authors studied estimating method on project duration in different modes based on GERT network diagram, and got the general steps of computing project duration. The method not only adapts to GERT network with reworking, also adapts to traditional determinate project network without reworking. Maturity of the method needs to be more verified in the future. ACKNOWLEDGEMENTS This research was supported by Open Foundation of Key Laboratory of Data Engineering and Knowledge Engineering, Ministry of Education P. R. China, Open Foundation of Key Laboratory of Information Management and Economy, Ministry of Information Industry, P. R. China under grants F0607-31, F0607-42, and Research Foundation of Renmin Univ. of China. REFERENCES 1. K. Kenzo and N. Nishiuchi, EJBicient Monte Carlo simulation method of GERT-type network for project management. Computers and industrial engineering. Volume 42, Number 2-4, pp.521-531, (2002). 2. E.H. John, Applications of simulation in project management, in Proceedings of the 11th conference on Winter simulation (San Diego, California, United States, 1979), pp.211-219. 3. K. Kim and J.M.D.L. Garza,. Evaluation of the resource-constrained critical path method algorithms. Journal of construction engineering and management-asce. Volume 131, Number 5, pp.522-532, (2005). 4. A. Azaron, H. Katagiri, and M. Sakawa, A multi-objective resource allocation problem in PERT networks, European Journal of Operational Research. Volume 172, Number 3, pp.838-854, (2006). 5. C.H. Cheng, Fuzzy repairable reliability-based on fuzzy GERT, Microelectronics and Reliability. Volume 36, pp.1557-1563, (1996). 6. E.R. Clayton and L.J. Moore, GERT vs. PERT, The Journal of System Management. Volume 22, pp.11-19, (1972). 7. E.D. David, PERT and simulation, in Proc. of the 10th conference on Winter simulation (Miami Beach, FL, 1978), pp.89-98.