UNIT:3 SOFTWARE PROJECT MANAGEMENT C \ ce @.0 ma\y “ e as 6RESPONSIBILITY OF SOFTWARE PROJECT MANAGER © Software project managers take the ore responsibility of steering a project to success. NL anager © The responsibilities and vi aie is large and varied. © The job resp tnd ei) manager ranges from ies inv) peti sy ike building up team morale to highly Iie ustomer presentations. © Manager also takes the responsibilities for project proposal writing, project cost estimation, scheduling, project staffing etc. ,© The activities of manager classify into two major types of responsibilities. © Project Planning: C e Project planning involves Shen NA | characteristics of the came ee the project activities ah es made. . rete the following activities. VE heduling o Staffing oRisk Management oMiscellaneous plans e© Project monitoring and control activities: e The project monitoring and control activities are undertaken once the development activities start. e The aim of the project monitoring and contr. re is to ensure that the ey ne iv F on veSKILLS REQUIRED IN PROJECT MANAGER o He must have theoretical knowledge of different project management techniques. C 0 A good decision making —t oe project manager. 400 o He should be oe e and has to determine and implem t need of the client, and capable of aga ig ee discussing the problems with customers. © He should have team leadership skill so the project which is divided into different persons are managed well and completed as per schedule. o He should have the experience in the related area of developing project.o He must be able to monitoring and tL the progress of the project. Evaluati fe f h A ° project. ing pertormance vt pets (J cp F oF aMETRICS FOR SIZE ESTIMATION oThe accurate estimation of the problem size cc fundamental to satisfactory estimation of REO project parameters such as oa ti eee completing the eels developing the Bove’ o The ag ize*is a measure of the problem apes in terms of the effort and time required to hig the product. cost for © Currently, two metrics are used to estimate size: e Lines of Code [LOC] e Function Point [FP]LINES OF CODE [LOC] © Simplest and most widely used metric to estimate the project size. WiC o It is popular method. cw 0 This metrics monk vy 6.8 oie by counting the TY struction in the developed pr o Dwring counting, the comments and blank lines should not be counted. © Determining the LOC count at the end of a project is very simple. But, accurate estimation of the LOC count the beginning of a project is very difficult.DISADVANTAGES OF LOC © LOC is language dependent. it focuses on coding TC only. © LOC gives numerical values ~ eae is depend on coding soe i er the size of specification. 2,20 ° 0 Lor giv axe with e quality and efficiency of code. sures lexical/textual complexity only. © Difficult to estimate LOC from problem description. So it is not useful for project planningFUNCTION POINT METRIC © The function point metric overcomes many of the shortcomings of the LOC metric. © The advantage of using the functio neni gat 1s e size of a that it can be used to easily een from the problem software produ btly > ee specific ° But i are size can be accurately determined only after the product has fully been developed. oln function point metric, the size of a software product is directly dependent on the number of different functions or features it supports.o Each function when invokes reads some input data and transforms it to the corresponding output data. Equation to count UFP ° UFP = (Number of inputs)*4 + (Number of outputs)*5 + (Number of Inquiries)*4 oust f Files)*10 + (Number of ree © Different Ane O t oy cope function a eal pps uts ber of Outputs e Number of Inquiries UFP (unadjusted e Number of Files e Number of Interfaces. eo Advantages of Function Point Itis not restricted to code. e Itis language independent. mic e Itis more accurate than csinannage OSS e We can have the mean Soe tl ecification that can’t be done with L 0 Di eu Function Point eMt ignores the quality of output. Itis fully oriented to traditional data processing system. e It doesn’t take into account the algorithmic complexity of software. ePROJECT ESTIMATION TECHNIQUES © Estimation of various parameters is a basic project planning activity. © The important project ee ewShx include: project size, Sey to develop the software, proje spo cost. mie eg ° Mi ‘ories of estimation techniques: “pnt estimation technique oHeuristic technique oAnalytical estimation techniqueEMPIRICAL ESTIMATION TECHNIQUES © This technique based on making an educated guess G the project parameters. oWhile using this ilar ede pr gests with development of similar paar o There Al tw. empirical techniques: appre rape phi estimation technique.EXPERT JUDGEMENT TECHNIQUE o In this technique, an expert makes an educated guess of C problem size after analyzing the problem. 4C o Usually, the expert estimates the c f\the hdd rent components that would ma ait te tem and then combines the iain AAthe ae modules to arrive at the over, ° nena is Sencue is subject to human errors and i ndividual bias. o Amore refined form of expert judgement is the estimation made by a group of experts.o Estimation by a group of experts minimizes factors such as individual oversight, lack of familiarity with a ti aspect of a project, personal bias and etc. OM "ne © However, the estimate made grape perts may still entire group of experts exhibit bias on iss ep be RF ieee to reasons such as_ political ° we the decision made by the group may be dominated by overly assertive member.DELPHI ESTIMATION TECHNIQUE © Delphi cost estimation technique tries to overcome some of the shortcomings of the expert judgment approach. ¢ Yr © Delphi estimation is carried out by a tea pene group of experts and a coordinator. % o In this approach, the Fey. Li les each estimator with a copy of e requirement specification(SRS) doc! np orm for recording his cost estimate. ator: o Est s complete their individual estimates anonymously and submit them to the coordinator.o The coordinator prepares the summary of the response of all the estimators. © Then the prepared summary information is aapgne ob estimators. o Based on this summary, ween © This process “oa rounds. eng NS cussion among the estimators is allowed We entire estimation process. o After ‘the completion of several iterations of estimators, the coordinator takes the responsibility of compiling the results and preparing the final estimate. EHEURISTIC TECHNIQUES © Heuristic techniques assume that the relationships among the different project parameters can be model suitable mathematical expressions. "WN 0 Once the basic ( seen re ne the other (dependent) e easily determined by substitutin: ue cy the basic parameters in the igh aor pression. ° ae models can be divided into the following two classes: single variable model and multivariable model.© Single variable estimation models provide a means a estimate the desired characteristics of a probl some previously estimated basic we eqaicovot the software product such as its rh Estimated pa o In above ee is a characteristics of the software ARNE, been estimated. ° wget: parameter is the dependent parameter to be estimated. ©The values of the constants cl and di are usually determined using data collected from past projects. ©}oA multivariable cost estimation model takes the following form: 1 o Estimated parameter = c, * ep, “4 +c, aN A Wa Sccts of the o Where ep, , ep, ,... are ts software ein c2,d1,d2,...are constants. o Multivari Bim ion oe are expected to give more ie imates compared to the single variable methods. © The COCOMO model is an example of this type of technique.COCOMO MODEL o COCOMO ( Constructive Cost Model) “~C proposed by Boehm. o It is regression based model mgt asst software cost estimation o Three types: & ove * Basie CGC HEME scone e Complete COCOMOo Basic COCOMO: development efforts using program size. It is quick and rough estimation te KC The basic COCOMO re) athys ah roximate estimate of the project parry) fort (E) = al * (KLOC)*? PM Tdev = b1 * (Effort)>? Months KLOC means Kilo Line of Code al,a2,b1,b2 are constants It is static, single values model that CuNtre Tdev means estimated time to develop the software in months Efforts means the total effort required to develop the sorwa product in Person Months.+ COCOMO applies to three classes of software projects: * Organic projects - "small" teams with "good" exper CG working with "less than rigid" malate ‘WN ° Semi-detached projects - mat Wa experience oe ¥e a ing ess than rigid Enel » Embed As qty within a set of "tight" Po 2 iso combination of organic and semi- Vics projects. (hardware, software, operational, ...)o Intermediate COCOMO model: « Intermediate COCOMO computes software develo CG effort as function of program size and a set o} etn that include subjective assessment, odes personnel and Project oye ti ve e This extension c vl "cost drivers", each with cyt ey eat - fware. utes attributes ° = Personnel attributes © Project attributes ¢ The Intermediate Cocomo formula now takes the form: E=a,KLoC)) (EAF)ADVANTAGE OF COCOMO MODEL o It is easy to use. o It is repeatable process. tC o It is versatile enough to suppo: questa and levels. Jt olt is hey api based on previous experi ° Tha be detailed documented.DISADVANTAGE OF COCOMO MODEL o COCOMO models are not accurate and not oC for scientific justification. cunie o It ignores software corngeey meh o It ignores personal “Gye, o It alsoi i meee ae issues. ° pony Der COMO are dependent on size Syl ayANALYTICAL ESTIMATION TECHNIQUE o Analytical estimation technique derive the required Te starting with certain basic a ph project. o Thus unlike empirical ay es h ay Wi ci analytical techniques do hevecsy We asis. 0 The seep exginigt nique. o This method is used to measure size, development effort, and development cost of software products. re science is an example of analytical© Halstead used a few primitive parameters to develop 6 expressions for over all program length, es volume, actual volume, language leygh development time. yt yew P ovSCHEDULING © Scheduling the project tasks is an important TC planning activity. Ie © It involves deciding whish tasks would eich o In order to schedule the ‘hon jot aus manager needs to xia ‘oe erro to complete the project. wn large tasks into small activities. Neck the dependency among different activities. a software project oEstablish the most likely estimates for the time durations necessary to complete the activities.© Allocate resources to activities. © Plan the starting and ending dates for various ew wees C o Determine the critical path. A “ara activities that determines Sie © project. P pO yrsWoOkK BREAKDOWN STRUCTURE o Work breakdown structure (WBS) is used to decompose c given task set recursively into small activities. ye o WBS provide a notation for order 2b needed to be carried out in ae faplem. © The root of the Bee roblem name. o Each at tree is broken down into smaller activities that de le the children of the node. ° wt activity is recursively decomposed into smaller sub- activities until at the leaf level. © The activities requires approximately two weeks to develop.ACTIVITY NETWORK oWBS _ representation a cewek Cb transformed into an activity representing the aoe eet nt (oy along with their int oy deye ce o Activity Gy ae graph using ‘Wifontorconnecting edges to represent ks and their planned sequence of completion, interdependence and interrelationship that must be accomplished to reach the project goals. eooAn activity network diagram helps to fin out the most efficient eC of, ye complete any Prolect olt enables aD t “reek project schedule pee ally pS Heal ere ount of time needed to complete the ‘oject. e The sequence in which tasks must be carried out e Which tasks can be carried out at the same time e Which are the critical tasks that you need to kee an eye on. Ooo In activity diagram, each activity is represented by a rectangular node and the duration fgg C activity is shown alongside each task wn oThe activity network Of displays interdependenci eyden sks through the use of boxes, apre S. ° Jt kn ing into a task box come from its predecessor tasks, which must be completed before the task can start. o Arrows pointing out of a task box go to its successor tasks, which cannot start until at least this task is complete.GANTT CHARTS o Gantt charts are mainly used to allocate resources to activities. iC o The resources allocated to sci HM staff, hardware and software. YY o Gantt charts ar i hy sources planning. oAG At is~a special type of bar chart where eagh' ar-trepresents an activity. o The bars are drawn along a time line. o The length of each bar is proportional to the duration of time planned for the corresponding activity.oThe Gantt Chart is used for software rod management, each bar consists of a oe. yea shaded part. o The shaded part of the Ae rg of time each task is esti o The eae ows Ast slack time, that is, the yy which a task must be finished.PERT CHARTS o PERT (Project Evaluation and Review damoysy (5 charts consists of a network of boxes and arro’ 4C o The boxes represent activities ang wp represent task dependencies o PERT chart repress gee Eee variations in the proje assuming a normal dist: ° mi a PERT chart instead of making a single estimate for each tasks, pessimistic likely, and optimistic estimates are made. o The boxes of PERT charts are usually annotated with the pessimistic, likely, and optimistic estimates for every task. e© Since all possible completion times between the minimum and maximum duration for a has to be considered, there is not wk critical paths, depending on t] eee s of the estimates for each Qet o This makes c j ting very ut oA AS ath in a PERT chart is shown by using thicker arrows. o Gantt charts representation of a project schedule is helpful in planning the utilization of resources, while PERT chart is useful for monitoring the timely progress of activities. alysis in PERT chartsRisk MANAGEMENT o Arisk is any anticipated unfavorable event or circumstance that can occur while a project .1e underway. WN o Risk management a gee ° Risk sacs ait an? mtainmentRIsK IDENTIFICATION o The project manger needs to anticipate the risks in re project as early as possible so that the imp aye risks can be minimized by e ayo management plans. o So, Early identificati ae o Risk Heng npi) ewhat ae to listing down you: ° Ake can be affected by a large variety of risks. o There are three main categories of risks which can affect a software project as follows: e Project Risk ° Technical Risk © ¢ Business Risko Project Risk: e Project risks concern various forms of bud, +e, schedule, personnel, resource, e c ant problems. e An important project seg ogi ea Cc e Since, software 4 it is very difficult to monitor and Wei. are project. 'e chnical Risk concern potential design, implementation, * interfacing, testing and maintenance problems. e Technical risks also include ambiguous specification, incomplete specification, changing specification etc. e Most technical risks occur due to the development tea insufficient knowledge about the product.o Business Risks: e This type of risks include risks of building an. CG excellent product that no one wants, los} ye or personnel commitments. v Cc pur por JRISK ASSESSMENT o The objective of risk assessment is to rank the risks in terms of their damage causing potential. o For risk assessment, first any risk pga two ways: e The likelihood of a va) e The conse chat gene et associated with that risk. EP ~ ese two factors, the priority of each risk a? computed: p=r*s o Where p is the priority with which the risk must be handled, r is probability of the risk becoming true, a: s is the severity of damage caused due to a becoming true.RISK CONTAINMENT o After all the identified risks of a project are iapaing plans must be made to first contain the i agnggin and the most likely risks. oc o Different risks SUN Te procedures. o There eee strategies to plan for risk a e Avoid the risk: oRisks can be avoided in several ways, such as discussing with the customer to change the requirements to reduce the scope of the work, givi incentives to the developers to avoid the isl manpower turnover. containmente Transfer the risk: oThis strategy involves getting the ris! phen developed by a third party, eK nen ‘over, and so on. ce ° Risk reductio: Ly “ap be pe ae ing ways to contain the damage For one if there is risk that some key personnel might leave, new recruitment may be planned. oThe most important thing to do in addressing technical risks is to build a prototype that tries out pieces of the technology that you are trying to use.