Network plan technology

The network plan technique uses network plans that describe a temporal and final chain of actions . It is used in particular in the scheduling of projects . Network plans, in turn, are the graphic illustrations of process chains. Since an activity can have several predecessor and, if necessary, subsequent activities, the image of a network of processes is created from which the name is derived.

Network planning uses concepts from graph theory . The network plan usually consists of a planar graph with nodes and edges as elements. The edges are directed and not cyclic in the network plan. There are two basic and dual variants

Process node network plan (VKN): In a process node network plan , processes are represented as nodes, the arrows show the arrangement and sequence relationships (example: Metra potential method , MPM) and

Activity arrow network plan (VPN): In an activity arrow network plan , activities are shown as arrows, the logical sequence is derived from the arrangement of the nodes (start / end of the activities) (example: method of the critical path , CPM).

Today, in construction and in plant engineering projects, the process node network plan is mainly used, in which the processes are represented by the nodes that are connected by edges. The basis for this form of representation was the Metra-Potential-Method . The edges describe different relationships: finish-start, start-start, finish-finish and start-finish relationships. There is also the display form of the task arrow network, also known as the critical path method .

More types

There are other variants of network plans:

Decision tree plan (ENP): Decision tree plans are based on the process arrow network plan technique and contain additional decision nodes as a logical element with optionally usable outputs and inputs. At the outputs, probability values can be assigned to the further routes. Such plans are usually implemented as pure tree structures.

Event node network plan (EKN): In an event node network plan , events (states) are shown as nodes and the time dependencies as arrows (example: Program evaluation and review technique , PERT).

Graphical Evaluation and Review Technique (GERT): Method from network analysis that is used in project management. It allows estimates to be taken into account, both from the network logic and from past process times. In the meantime, however, GERT is no longer regarded as generally applicable and is used very rarely. This is why it is no longer recommended by the American organization Project Management Institute (PMI).

Application recommendation

Activity arrow networks, e.g. B. CPM, can be used when the project has simple relationships that rarely need to be changed. CPM network plans are less suitable for cost and resource planning. Activity node networks, e.g. B. MPM, have the advantage that the process nodes can be assigned a lot of different information. Another advantage is that they can be changed relatively quickly. Event node networks such as PERT should be used in projects in which the processes cannot be precisely predefined in terms of time or structure (stochastic network method).

Because of their relatively easy handling, CPM and MPM are the most common.

Processing of network plans

In currently available software for network planning, the original methods are combined and can no longer be separated. Network plans are set up incrementally and iteratively using suitable software. The graphic representations are helpful to keep the overview. However, the detailed processing is usually carried out for each individual process, which offers corresponding entry fields in a user interface for each process and for each subsequent condition and for time condition.

A parser checks that there are no logical errors (no loops , no negative durations, etc.) as an automatic process when recording and calculating appointments. If required, the refinement takes place by breaking down complex processes into further individual processes and by introducing further relationships between the processes. An essential task of the processing of network plans is the updating of the dates and the processing status for ongoing and completed processes as well as for changed provisions and newly recognized dependencies.

Goals of the network plan technique

May be a basis for network planning project structure plan to be.

The goal of network planning is the planning of the logical relationships between the processes and the timing of the processes. The network plan forms the basis for the creation of communication tools such as B. Milestones , bar chart or networked bar chart .

The use of network plan technology should answer four important questions:

How long will the whole project take? What are the risks?
What critical activities can delay the entire project if they are not completed on time?
Is the project on schedule, will it be finished sooner or later?
If it is to be finished sooner, what is best to do, how can it be accelerated with the least cost?

The concept of network planning is based on the experience that a few activities, which form the longest path through the network, determine the course of the entire project. If these critical activities (= critical path) are recognized early, countermeasures can be taken early on. Management can concentrate on the critical activities. Non-critical activities can be rescheduled without affecting the entire project.

Purposes of network planning

The logical connections of a project from the beginning to the end can be clearly displayed.
A schedule can be developed for all project processes with the help of the network plan.
A critical path and resource bottlenecks that can jeopardize the deadline can be easily identified.
Network plans form the basis for ongoing project control and deadline monitoring .

In the network plan technique, there are four sub-tasks:

Capacity planning: planning the required means of production
Cost planning
Structural planning
Time planning : Activities are assigned time values ( time window ).

The technique is called network plan technique because the graphical representation of circles and arrows in these plans looks like a network in projects with many processes. Before the widespread use of computers, network plans were still drawn with a pencil or they were made with the help of painted steel sheets and magnets. These had the advantage that changes to the project could quickly be incorporated into the project planning for the circumstances at the time, without having to draw the project plan again.

Basic concepts

Definition in standardization

According to DIN 69900-1, the term network plan technology includes “all methods for the analysis , description, planning , control and monitoring of processes based on graph theory , whereby time, costs, resources and resources can be taken into account. A network plan is the graphic or tabular representation of processes and their dependencies ”.

process

In the context of network planning, a process is a delimited unit of work that begins at a certain point in time and ends at a certain later point in time. In more general terms: "A process is a process element that describes a certain event."

In network planning, the term “process” is used, in contrast to the terminology used in project management, which prefers the term “work package” as a planning unit.

Tasks are usually parts of the project process; in the exceptional case of the waiting process, however, there is no sequence. A process can be linked to other processes: For example, a "putting on socks" process would have to be completed before a "putting on shoes" process can be started. These dependencies are discussed in detail below.

Such a process has an essential property, its duration . The task of network planning technology is to determine when the respective processes take place , taking into account the duration of the individual processes and taking into account their dependencies . Depending on requirements, the computing process begins either with the start processes and, based on these, sets the earliest possible start date for the subsequent processes (forward planning), or with the last processes in the network (which no longer have a successor), and then sets the latest completion dates of the respective upstream processes (backward planning). By combining both methods, based on a defined start and a defined end date, the following four other important properties result for each process in addition to the duration:

F rühester A nfangs z eitpunkt (FAZ) (from forward planning)
F rühester E nd z eitpunkt (FEZ) (from forward planning and respective duration)
S pätester E nd z eitpunkt (SEZ) (from reverse engineering)
S pätester A nfangs z eitpunkt (SAZ) (from reverse planning and respective duration)

Dates of an activity : ${\ displaystyle V_ {ij}}$

${\ displaystyle FAZ_ {ij}}$ : Earliest start time of an activity ${\ displaystyle V_ {ij} = FEZ_ {i}}$

${\ displaystyle FEZ_ {ij}}$ : Earliest end time of a task ${\ displaystyle V_ {ij} = FAZ_ {ij} + D_ {ij}}$

${\ displaystyle SEZ_ {ij}}$ : Latest end time of a task (in compliance with the project end date ) ${\ displaystyle V_ {ij} = SAZ_ {j}}$

${\ displaystyle SAZ_ {ij}}$ : Latest start time of an activity (in compliance with the project end date) ${\ displaystyle V_ {ij} = SEZ_ {ij} -D_ {ij}}$

Edge slip

(Short form KS) denotes the leeway that an arrangement relationship (AOB = networking according to DIN 69900) has to the subsequent process in relation to the earliest position.
The free buffer (FP for short) of a transaction is calculated from the smallest edge slip of all subsequent AOBs.

The following graphic display forms are used as standard in the network plan:

KS = 0 ->> double arrow = AOB defining the date = AOB moves the successor
KS> 0 -> Single arrow = AOB has room for maneuver / slack = AOB goes into "emptiness"
KS <0 ->>>triple arrow = AOB is stuck around the negative value of the edge slip
(for example, because the successor already has a fixed start date and can no longer be postponed.)

Representation of the edge slippage in the network plan

In the case of project management software products that calculate according to the Metra Potential method , the chains that determine the dates can be tracked via the edge slip.

Examples:

In the case of PRIMAVERA, the AOB that determines the date is marked with *
With ACOS PM software, the edge slip is even displayed with a numerical value and can also be selected

Buffer time

The buffer time is a time margin for the execution of a process, so-called time reserves. This leeway can be used by postponing the process and / or by lengthening (stretching) the duration of the process.

Four types of buffer time can then be determined in the network plan from the information provided by several specifications :

The total buffer of a task is the period of time that a task can be moved from its earliest start (or duration) without endangering the end of the project ${\ displaystyle GP_ {i}}$

{\ displaystyle GP_ {i} = SAZ_ {i} -FAZ_ {i}}

An operation is critical when its total buffer is zero.

The free buffer is the time that does not endanger the earliest possible start or end of the successor. (Formal: All follow-up processes can be carried out in their earliest position). It can only arise if at least two completed processes encounter the same successor. In the case of a “normal sequence” (end - beginning), it is calculated by forming the difference between the earliest end of the process under consideration and the earliest start of its successor. With a start sequence (start - start) the earliest start dates and with an end sequence (finish - finish) the earliest end dates of the activities are compared. Example normal sequence:

{\ displaystyle FP_ {i} = min_ {j \ in S (i)} (FAZ_ {j} -b_ {ij}) - FEZ_ {i}}

FP = FAZ (Nachf.) - FEZ

{\ displaystyle b_ {ij}}

= minimum time interval between process i and process j

{\ displaystyle S (i)}

= Set of successors to activity i

The free backward buffer is the maximum amount of time by which the task can be postponed starting from its latest possible start time, but on the condition that all previous tasks are on the latest possible date. It can only come into being if any previous predecessor has at least two successors. The following calculation formula applies under the condition that all links between processes i and i-1 of type EA = 0.

{\ displaystyle FRP_ {i, j}}

= -

{\ displaystyle SAZ_ {j}}

{\ displaystyle SEZ_ {i}}

The independent buffer is the maximum amount of time that the task can be postponed if all previous tasks should end at the latest possible date and all subsequent tasks should start at the earliest possible date. The use of the independent buffer therefore has no effect on the position of the predecessor and successor. The following calculation formula applies under the condition that all links between processes i and i-1 are of type EA = 0.

{\ displaystyle UP_ {i, j}}

= max {0, - - }

{\ displaystyle FAZ_ {j}}

{\ displaystyle SEZ_ {i}}

{\ displaystyle D_ {i, j}}

The total buffer is primarily of practical importance.

${\ displaystyle UP = FRP + FP-GP}$

${\ displaystyle UP = (FAZ-maxSEZ (i-1)) + (minFAZ (i + 1) -FEZ) - (SAZ-FAZ)}$

${\ displaystyle UP = FAZ-FEZ + minFAZ (i + 1) -maxSEZ (i-1)}$

Critical path

The critical path , also known as the critical path , is the chain of those processes whose timing changes the end date of the network plan. It is determined in a network plan by a chain of individual activities, the total buffer time of which is zero.

The activities that are on the critical path determine the total project duration. All other activities can be postponed or extended within their buffer time without changing the overall project duration. The total buffer for the critical path is zero.

More terms

Event : An event is the occurrence of a defined state in the process. Events represent times at which certain sub-processes are finished or others should begin; the beginning and the end of a process are consequently events. An event does not extend in time. They form the basis for milestone planning.
Relationships : Relationships identify the logical dependencies between events or processes. There are four possibilities for the sequence of two processes A and B:
- End-start : B can be started as soon as A has ended (EA relationship or normal sequence).
- Start-start : B can be started as soon as A has been started (AA relationship or start sequence).
- Start-end : B can be ended as soon as A has been started (AE relationship or jump sequence).
- End-end : B can be ended as soon as A has been ended (EE relationship or end sequence).
Process structure : The process structure of a network plan is represented by the totality of the relationships

Plan representation

The network diagram itself only provides an overview and is complex to produce. It does not serve as a working document and is mostly only used when a network plan is to be revised manually.

Example of a network plan

Working with network plans can be divided into:

Draft as the breakdown of the project task into processes or events, taking into account logical and causal relationships. The design is the most important and also the most difficult part of the work, because it is the only thing that matters whether the result of the planning makes sense or not.
Time analysis in the form of an estimate / calculation of the process duration (or duration between two events). A good estimate of the times is the second most important and also difficult task. The gain in knowledge from design and time estimation is much greater than the subsequent calculation of the network plan with the determination of the critical path and the time reserves.
Project monitoring through corrections to the network plan and monitoring of the project progress.

With a comprehensive application, cost and resource analyzes are also possible.

Before the actual network plan is created, the dependencies of the individual processes and their duration are often recorded in a table.

The network plan is then created, with each process being graphically recorded as boxes (network plan nodes). A description as well as the duration, earliest start, latest start, earliest end and latest end are noted at defined positions. In addition, fields are reserved for the total buffer and the free buffer.

D.	Duration
GP	Total buffer
FP	free buffer
FB	earliest start
SB	latest start
FE	earliest end
SE	latest end

The individual boxes per activity are connected by arrows, which show the dependencies between the activities. Since no cyclical linkages are allowed, the individual processes can be arranged in their necessary chronological order from left to right, with parallel paths below one another.

Simple example of a network plan

To calculate the buffer times, the duration of the earliest start of the first event is added. The result is at the same time the earliest end time of the current task and the earliest start of the next. After you have arrived at the last process with this forward calculation, you start from there with the backward calculation of the latest project end specified by the client, as the latest end. The difference between the earliest and latest start results in the total buffer.

You can then take the following information from the example network diagram shown here:

The project ends after six days at the earliest.
The critical path includes operations AA and CC .
Process BB can also only be started after a buffer time of two days without endangering the project.

Advantages of the network plan technique

The network plan is a very understandable, clear and meaningful representation because it provides an excellent overview of the totality of the sub-processes of a project and their mutual dependencies. Due to the mapping of the entire project in the network plan, it is necessary to think through this entire project and not to limit oneself to a single sub-project. The network plan itself can be captured quickly and therefore easily updated and enables a relatively exact prediction of important intermediate dates and the final date. Furthermore, time bottlenecks (critical path) and buffer times are easily recognizable in the network plan, which can be assigned by project management software, for example, in different colors for processes on the critical path.

In contrast to planning with bar charts , a distinction can be made in network planning between process planning and scheduling. The ability to display the logical dependencies between processes enables a schedule to be developed independently of scheduling assumptions. In short it can be said:

Network plans form an understandable, clear and meaningful representation of the entire project process.
They can be recorded quickly and easily updated (provided the project data is processed electronically).
Critical processes and bottlenecks are easy to identify.
When creating them, it is necessary to think through the entire project.
Much of the available project management software supports network planning.
Network plans force you to systematically think through the project context.
Are a flexible information medium to ensure the exchange of data between project management and the executing and superior departments.
Show where there are time reserves (buffers), where they are missing and where acceleration measures are unavoidable.
Allow the secure termination of sub-projects / work packages.
Enable the proper control of the project process with regard to costs, capacities, deadlines.

An example of a network plan-oriented project management software is Micro Planner .

Disadvantages of the network plan technique

The graphical representation, the network plan, is often used for larger projects. In some cases, smaller projects can be presented in a more user-friendly way using a representation derived from this, the bar chart ( Gantt diagram ). Alternatives to network plan technology would be the Gantt or PLANNET technology mentioned above (a further development of Gantt technology).
The increasing level of detail in a network plan is directly linked to increasing control and revision effort in order to avoid errors due to deviations between the planned and actual state of a project.
If the network plan is structured too abstractly or in a way that is not practical and therefore work steps are described in an incomprehensible manner, the probability is high that it will not be understood by the users.

Secondary literature

Awni Al-Ani: Practice of project planning with network plan technology . Dr. Otto Schmidt KG, Cologne 1971, ISBN 978-3-504-52000-7 .
Michael Sauer: Operations Research compact , 1st edition. Oldenbourg, Munich 2009, ISBN 978-3-486-59082-1 .
Klaus J. Bechler, Dietmar Lange: DIN standards in project management . Beuth Verlag, sales number 16005

Web links

Wiktionary: Netzplanechnik - explanations of meanings, word origins, synonyms, translations

Individual evidence

↑ Activity node network plan
↑ Task arrow network plan
↑ Decision tree analysis
↑ Event node network plan
↑ Graphical Evaluation and Review Technique, 152 pages (PDF; 3.4 MB)
↑ DIN 69900 . Project management - network technology; Descriptions and terms.
^ Erhard Motzel: Project Management Lexicon. 2nd edition, Weinheim 2010: Wiley-VCH, p. 122. ISBN 978-3-527-50471-8

[1] Activity node network plan

[2] Task arrow network plan

[3] Decision tree analysis

[4] Event node network plan

[5] Graphical Evaluation and Review Technique, 152 pages (PDF; 3.4 MB)

[6] DIN 69900 . Project management - network technology; Descriptions and terms.

[7] Erhard Motzel: Project Management Lexicon. 2nd edition, Weinheim 2010: Wiley-VCH, p. 122. ISBN 978-3-527-50471-8