White Pages | Index | Introduction | Basics | Scheduling the Project | More Advanced Topics

Introduction
Before you turn the machine on
Networks
Network Logic
Network Structure
Activities
Durations
Constructing the Network
Network Elements

Hornet uses a project management technique known as Precedence Networking or CPA.

Precedence networking is an invaluable planning aid in that it enforces systematic analysis, making it easier to visually represent complex projects and allowing you to identify many problems at a stage when they may be solved or avoided.

Hornet combines the power of the computer with precedence networking. This offers the power to update information quickly, adding flexibility to the planning process and enhancing the monitoring process. Hornet makes it possible to assess progress realistically on a weekly or daily basis.

The precedence networking technique involves an understanding of networks and network logic.

In order to plan a project, you have to decide on its network structure and its network logic; and determine the activities and durations. You must do all of this before you can begin to use Hornet. The computer cannot perform these tasks for you. Therefore, you have to:

• Decide whether the project needs to be divided into subnetworks and, if so, decide what they are.
• Choose appropriate activities and give them realistic durations.
• Put the activities into a logical sequence, listing each activity's precedents (those activities which cannot start before others). The best way to do this is to get out a pencil and actually plan the project on paper. Nevertheless, some planners prefer to enter data straight into the computer system, specifying the activities, durations and logic as they proceed. You can do this with Hornet if you wish.

Precedence networking is one of many network analysis techniques based on the concept that any project can be divided into a number of activities, or tasks, which must be accomplished before the project is finished.

A network is made up of activities, their sequence and inter-relationship, and it can be represented diagrammatically. Time moves from left to right, although there is no scale. Hence a network can be seen as a logical grouping of activities within a project and any project can have one or more networks.

Each activity is represented by a rectangular box called a node and the relationships between the activities are represented by lines.

Network logic describes the order in which activities happen and specifies the dependencies, that is, which activities are dependent on other activities.

The activities in a project are interrelated in a way that depends on the circumstances.

For example, if we are building a house we lay the foundations before building the walls, and build the walls before putting on the roof. This is the only order in which to do the job.

Precedence networking is so called because it is based on the systematic listing of each activity's precedents; that, is, the activities that must logically precede each activity. 'Laying foundations' is a precedent of 'Building walls' because building the walls cannot start until laying the foundations is finished.

By describing all the activities in a network in this way we build up a picture of the relationships between them, which becomes the basis for a plan. Then, using the computer, we look at all possible solutions until the most effective becomes apparent.

There are three types of logical activity relationships used in precedence networking: finish-to-start, start-to-start and finish-to-finish. These are precedent links that describe the relationship between an activity and those that precede it.

The network logic may be effected by logic delay and internetwork links.

For many projects, one network is sufficient. However, other projects divide naturally into broad segments.

For example, if you were building a hydroelectric power station, building the power station could be one segment, while building the reservoir could be another, and building the spillway could be a third.

Instead of putting these segments into one network, you can put them into separate subnetworks, these are controlled by a governing network which is network zero or the MASTER network of the project.

In Hornet, the name and number of the subnetwork is always the same as that of its master activity in network zero. Thus, network 100 is called 'Build power station' and activity 100 in network zero is also called 'Build Power Station'. Hornet automatically links each subnetwork with network zero.

You may also wish to link subnetworks on the same level. Suppose, for example, you have only one large crane, which you need both for the power station and the reservoir. You would insert an internetwork link between the two subnetworks that would ensure that the crane's work on the reservoir is not scheduled to begin before its work on the power station is finished.

As you can see, the terms' network and subnetwork are similar in meaning. Any subnetwork is a network in its own right, however network zero is never a subnetwork.

Activities must do two things: divide the project into a linked pattern of essential tasks; and provide appropriate information to the people who need it. Your job is to create a network, and a series of subnetworks if necessary, which satisfies these needs.

Both network zero and the subnetworks contain activities. In order to make the plan successful, the activities have to be appropriate.

Consider the future uses of the plan. Before implementation, a plan is a map of the future. It needs to be reasonably general.

An activity such as 'Have secretary type specifications' is trivial: there is no point in planning typing unless typists are a very scarce resource. On the other hand, an activity such as 'Build power station' is too general. It needs to be broken down into component bits.

Therefore, the function of subnetworks before implementation is to identify the essential elements of the project and to establish their relationships. Once the project is underway, the main function of subnetworks is to provide benchmarks against which the actual progress of the project can be measured.

With Hornet you create the reports you and other key people in the project need. These reports are based on activities. Senior management will require reports that only show the overall picture. Supervisors or foremen will want to know exactly what their own teams are supposed to do, and when.

Hornet allows you to produce reports that show certain groups of activities, for example all the activities that a particular team will carry out. By setting up a management code system, you can produce highly selective customised reports.

In addition to determining the network logic, subnetwork structure and the activities, we have to consider one further item if we want to make a successful plan. We need to know how long the project is going to take.

Broadly, we work this out by estimating the duration of each activity and adding the durations up to find the total. As the accuracy of this total is dependent on the accuracy of the individual durations, it is important to estimate them realistically.

A large part of the skill of project management lies in making realistic estimates of activity durations.

The methods for estimating activity durations are:-

• look at how long similar activities have taken in the past
• consult people who actually perform or are familiar with the task

· educated guesswork based on experience and intuition.

However, because you are going to use Hornet to plan your project, it is not necessary to spend too much time worrying about durations at this point. One of the great advantages of computerised project management is that you can easily change data such as durations while planning.

Drawing the network enforces systematic thinking, helping you to define it more effectively.

A network can be seen as a statement of what work is going to be done to achieve a given end.

If several project managers were asked to plan the same project, they would probably come up with quite different but equally effective ways of tackling it. By the same token, their subnetworks would all be different. There is no 'right' or 'wrong' ways of networking to produce solutions, only creative ways of tackling problems.

By visually representing project logic, you can more easily identify inefficient or unnecessary procedures and logical inconsistencies in your plan.

Before you use Hornet you should try to ensure that your project plan and network diagrams accurately reflect reality. This will reduce the likelihood that you will encounter logical inconsistencies and other problems later on.

A network has two major elements: nodes, and lines that indicate the network logic.

Nodes

Nodes are used in networks to represent activities. This is a node of the type recommended by the British Standards Institute.

This shows the activity's name and identifier; its duration; its float; and the earliest and latest dates it can start and finish. However, as Hornet calculates float and the start and finish dates for you, the nodes in your network diagram need only show the activity name or number, estimated duration, and if you wish, description.

Although a network in Hornet can contain many activities, we recommend that unless you are planning a very large project a network should not contain more than 100 activities. If you are drawing the network out on paper before putting it into Hornet, you should be able to fit the whole network onto one piece of A3 paper. This may not sound like many activities in a network, but if you put too many activities in a network you may find that your plan is too detailed to be really useful.

Remember that any activity in network zero can be an entry point to a subnetwork and you can put each subnetwork on a separate sheet of paper.

In order to make it easier to understand the logic of the networks when reading the data in reports we recommend that you number the activities so that the connection between them is apparent. For example, preceding activities might have lower numbers than succeeding activities.

Lines

The lines in your network diagram represent the relationships between the activities. The relationships between activities may be:-

• Finish-to-start
• Start-to-start
• Finish-to-finish

Finish-to-start (FS) link

The first activity must finish before the second one can start. (The term link describes the relationship between the activities). FS links are represented in network diagrams as shown below.

Hornet assumes that there is no delay between an activity and its precedent. If you have a situation in which the succeeding activity does not start immediately after its precedent, you would use an FS link with a logic delay.

Start-to-start (SS) link

These activities can start at the same time. SS, or lead links are represented in network diagrams as shown below.

For example, suppose your activities were ‘Paint walls’ and ‘Paint roof’. You do not have to finish painting the walls before we start painting the roof, or vice versa. These activities can start at the same time. You would put a start-to-start (or SS) link between the activities 'Paint walls' and 'Paint roof'.

The classic example of an SS link is laying a pipe. Before you can lay a pipe, you have to dig the trench to put it in. It would not make much sense to put finish-to-start links between the operations: you do not have to dig the entire trench before you start laying the pipe. On the contrary, you can begin laying the pipe very soon after you have started digging.

You would therefore use a start-to-start link between the two, but introduce a logic delay between them. The logic delay would represent the amount of time you would have to wait for the first activity to get going before you could start the second.

This diagram indicates that you expect to wait one day before beginning to lay the pipe.

Finish-to-finish (FF) link

This link represents a situation in which two activities finish at the same time. FF, or lag, links are represented in network diagrams as shown below.

For example, suppose you were laying a pipe in a trench. As you could not finish laying it until you had finished digging the trench to put it in, you would put a finish-to-finish (or FF) link between the activities 'Dig trench' and 'Lay pipe'.

When activities are connected by finish-to-finish links there is no constraint on their starts. This may cause an activity to display preceding float.

Logic Delay

A delay between two activities connected by a finish-to-start link can be represented like this (where d is the delay):

For example, suppose you are painting a radiator. If you apply two coats of paint, you have to let the first dry before you apply the second.

You would therefore put an FS link between the activities 'First Coat' and 'Second Coat', but put a logic delay between them. The logic delay would be the amount of time it took for the first coat to dry.

Logic Delay Calendar

Every activity in a project operates according to a calendar. These calendars affect the links between activities and thus affect the timing of the logic delay. The logic delay may fall in a different place depending on the calendar that governs it.