Reference: Sommerville, Software Engineering, 10 ed., Chapter 23

 

Plans are useless, but planning is indispensable
General Dwight D. Eisenhower

Project planning involves breaking down the work into parts and assigning these to project team members, anticipate problems that might arise and prepare tentative solutions to those problems. The project plan, which is created at the start of a project, is used to communicate how the work will be done to the project team and customers, and to help assess progress on the project.

At the proposal stage, you are bidding for a contract to develop or provide a software system. Planning may be necessary with only outline software requirements. The aim of planning at this stage is to provide information that will be used in setting a price for the system to customers. Project pricing involves estimating how much the software will cost to develop, taking factors such as staff costs, hardware costs, software costs, etc. into account.

During the project startup phase, you have to plan who will work on the project, how the project will be broken down into increments, how resources will be allocated across your company, etc. At this stage, you know more about the system requirements but do not have design or implementation information. Create a plan with enough detail to make decisions about the project budget and staffing. This plan is the basis for project resource allocation. The startup plan should also define project monitoring mechanisms. A startup plan is still needed for agile development to allow resources to be allocated to the project.

Periodically throughout the project, you need to modify your plan in the light of experience gained and information from monitoring the progress of the work. The project plan should be regularly amended as the project progresses and you know more about the software and its development. The project schedule, cost-estimate and risks have to be regularly revised.

Estimates are made to discover the cost, to the developer, of producing a software system. You take into account, hardware, software, travel, training and effort costs. There is not a simple relationship between the development cost and the price charged to the customer. Broader organizational, economic, political and business considerations influence the price charged.

Factors affecting software pricing:

Contractual terms

A customer may be willing to allow the developer to retain ownership of the source code and reuse it in other projects. The price charged may then be less than if the software source code is handed over to the customer.

Cost estimate uncertainty

If an organization is unsure of its cost estimate, it may increase its price by a contingency over and above its normal profit.

Financial health

Developers in financial difficulty may lower their price to gain a contract. It is better to make a smaller than normal profit or break even than to go out of business. Cash flow is more important than profit in difficult economic times.

Market opportunity

A development organization may quote a low price because it wishes to move into a new segment of the software market. Accepting a low profit on one project may give the organization the opportunity to make a greater profit later. The experience gained may also help it develop new products.

Requirements volatility

If the requirements are likely to change, an organization may lower its price to win a contract. After the contract is awarded, high prices can be charged for changes to the requirements.

Pricing strategies:

Under pricing

A company may underprice a system in order to gain a contract that allows them to retain staff for future opportunities. A company may underprice a system to gain access to a new market area.

Increased pricing

The price may be increased when a buyer wishes a fixed-price contract and so the seller increases the price to allow for unexpected risks.

The software is priced according to what the software developer believes the buyer is willing to pay. If this is less that the development costs, the software functionality may be reduced accordingly with a view to extra functionality being added in a later release. Additional costs may be added as the requirements change and these may be priced at a higher level to make up the shortfall in the original price.

Plan-driven or plan-based development is an approach to software engineering where the development process is planned in detail. Plan-driven development is based on engineering project management techniques and is the 'traditional' way of managing large software development projects. A project plan is created that records the work to be done, who will do it, the development schedule and the work products. Managers use the plan to support project decision making and as a way of measuring progress. The arguments in favor of a plan-driven approach are that early planning allows organizational issues (availability of staff, other projects, etc.) to be closely taken into account, and that potential problems and dependencies are discovered before the project starts, rather than once the project is underway. The principal argument against plan-driven development is that many early decisions have to be revised because of changes to the environment in which the software is to be developed and used.

In a plan-driven development project, a project plan sets out the resources available to the project, the work breakdown and a schedule for carrying out the work. Plan sections include:

• Introduction
• Project organization
• Risk analysis
• Hardware and software resource requirements
• Work breakdown
• Project schedule
• Monitoring and reporting mechanisms

Project planning is an iterative process that starts when you create an initial project plan during the project startup phase. Plan changes are inevitable. As more information about the system and the project team becomes available during the project, you should regularly revise the plan to reflect requirements, schedule and risk changes. Changing business goals also leads to changes in project plans. As business goals change, this could affect all projects, which may then have to be re-planned.

Planning assumptions: You should make realistic rather than optimistic assumptions when you are defining a project plan. Problems of some description always arise during a project, and these lead to project delays. Your initial assumptions and scheduling should therefore take unexpected problems into account. You should include contingency in your plan so that if things go wrong, then your delivery schedule is not seriously disrupted.

If there are serious problems with the development work that are likely to lead to significant delays, you need to initiate risk mitigation actions to reduce the risks of project failure. In conjunction with these actions, you also have to re-plan the project. This may involve renegotiating the project constraints and deliverables with the customer. A new schedule of when work should be completed also has to be established and agreed with the customer.

Project scheduling is the process of deciding how the work in a project will be organized as separate tasks, and when and how these tasks will be executed. You estimate the calendar time needed to complete each task, the effort required and who will work on the tasks that have been identified. You also have to estimate the resources needed to complete each task, such as the disk space required on a server, the time required on specialized hardware, such as a simulator, and what the travel budget will be. Project scheduling activities:

• Split project into tasks and estimate time and resources required to complete each task.
• Organize tasks concurrently to make optimal use of workforce.
• Minimize task dependencies to avoid delays caused by one task waiting for another to complete.
• Dependent on project managers intuition and experience.

Scheduling problems:

• Estimating the difficulty of problems and hence the cost of developing a solution is hard.
• Productivity is not proportional to the number of people working on a task.
• Adding people to a late project makes it later because of communication overheads.
• The unexpected always happens. Always allow contingency in planning.

Graphical notations are normally used to illustrate the project schedule. These show the project breakdown into tasks. Tasks should not be too small. They should take about a week or two. Bar charts (Gantt charts) are the most commonly used representation for project schedules. They show the schedule as activities or resources against time. They may also show task dependencies.

Milestones are points in the schedule against which you can assess progress, for example, the handover of the system for testing. Deliverables are work products that are delivered to the customer, e.g. a requirements document for the system.

Agile methods of software development are iterative approaches where the software is developed and delivered to customers in increments. Unlike plan-driven approaches, the functionality of these increments is not planned in advance but is decided during the development. The decision on what to include in an increment depends on progress and on the customer's priorities. The customer's priorities and requirements change so it makes sense to have a flexible plan that can accommodate these changes.

Agile planning stages:

• Release planning looks ahead for several months and decides on the features that should be included in a release of a system.
• Iteration planning has a shorter term outlook, and focuses on planning the next increment of a system. This is typically 2-4 weeks of work for the team.

Approaches to agile planning:

• Based on managing a project backlog (things to be done) with daily reviews of progress and problems.
• The planning game developed originally as part of Extreme Programming (XP). Dependent on user stories as a measure of progress in the project.

The planning game is based on user stories that reflect the features that should be included in the system. The project team read and discuss the stories and rank them in order of the amount of time they think it will take to implement the story. Stories are assigned effort points reflecting their size and difficulty of implementation. The number of effort points implemented per day is measured giving an estimate of the team's velocity. This allows the total effort required to implement the system to be estimated.

Release planning involves selecting and refining the stories that will reflect the features to be implemented in a release of a system and the order in which the stories should be implemented. Stories to be implemented in each iteration are chosen, with the number of stories reflecting the time to deliver an iteration (usually 2 or 3 weeks). The team's velocity is used to guide the choice of stories so that they can be delivered within an iteration.

During the task planning stage, the developers break down stories into development tasks. A development task should take 4-16 hours. All of the tasks that must be completed to implement all of the stories in that iteration are listed. The individual developers then sign up for the specific tasks that they will implement. Benefits of this approach are that the whole team gets an overview of the tasks to be completed in an iteration, and that the developers have a sense of ownership in these tasks and this is likely to motivate them to complete the task.

A software increment is always delivered at the end of each project iteration. If the features to be included in the increment cannot be completed in the time allowed, the scope of the work is reduced. The delivery schedule is never extended.

Agile planning is reliant on customer involvement and availability. This can be difficult to arrange, as customer representatives sometimes have to prioritize other work and are not available for the planning game. Furthermore, some customers may be more familiar with traditional project plans and may find it difficult to engage in an agile planning process.

Agile planning works well with small, stable development teams that can get together and discuss the stories to be implemented. However, where teams are large and/or geographically distributed, or when team membership changes frequently, it is practically impossible for everyone to be involved in the collaborative planning that is essential for agile project management.

Organizations need to make software effort and cost estimates. There are two types of technique that can be used to do this:

Experience-based techniques

The estimate of future effort requirements is based on the manager's experience of past projects and the application domain. Essentially, the manager makes an informed judgment of what the effort requirements are likely to be.

Algorithmic cost modeling

In this approach, a formulaic approach is used to compute the project effort based on estimates of product attributes, such as size, and process characteristics, such as experience of staff involved.

Experience-based techniques rely on judgments based on experience of past projects and the effort expended in these projects on software development activities. Typically, you identify the deliverables to be produced in a project and the different software components or systems that are to be developed. You document these in a spreadsheet, estimate them individually and compute the total effort required. It usually helps to get a group of people involved in the effort estimation and to ask each member of the group to explain their estimate. The difficulty with experience-based techniques is that a new software project may not have much in common with previous projects. If you have not worked with these techniques, your previous experience may not help you to estimate the effort required, making it more difficult to produce accurate costs and schedule estimates.

With algorithmic cost modeling, Cost is estimated as a mathematical function of product, project and process attributes whose values are estimated by project managers:
Effort = A * Size^B * M
A is an organization-dependent constant, B reflects the disproportionate effort for large projects and M is a multiplier reflecting product, process and people attributes. The most commonly used product attribute for cost estimation is code size. Most models are similar but they use different values for A, B and M. The size of a software system can only be known accurately when it is finished. Several factors influence the final size: use of reused systems and components; programming language; and distribution of system. As the development process progresses then the size estimate becomes more accurate. The estimates of the factors contributing to B and M are subjective and vary according to the judgment of the estimator. Algorithmic cost models are a systematic way to estimate the effort required to develop a system. However, these models are complex and difficult to use. There are many attributes and considerable scope for uncertainty in estimating their values. This complexity means that the practical application of algorithmic cost modeling has been limited to a relatively small number of large companies, mostly working in defense and aerospace systems engineering.

COCOMO (Constructive Cost Modeling) cost modeling is an empirical model based on project experience. It is a well-documented, 'independent' model which is not tied to a specific software vendor. Long history from initial version published in 1981 (COCOMO-81) through various instantiations to COCOMO 2. COCOMO 2 takes into account different approaches to software development, reuse, etc. COCOMO 2 incorporates a range of sub-models that produce increasingly detailed software estimates. The sub-models in COCOMO 2 are:

• Application composition model used when software is composed from existing parts.
• Early design model used when requirements are available but design has not yet started.
• Reuse model used to compute the effort of integrating reusable components.
• Post-architecture model used once the system architecture has been designed and more information about the system is available.