Schedule Integrity is your go-to feature to see what nPlan has read from your schedule file and what are the top integrity issues that may affect your results
You can access Schedule Integrity by clicking on the button present in the project home page. Then, you will be taken to the report which contains the following. The Schedule Integrity report contains DCMA-14 checks plus some checks that are specific to nPlan’s forecasting engine.
- Summary — Contains general information about the activities that have been read from the file and the target milestone that you have selected for forecast:
- Total planned duration — Is the date difference between the target milestone and the start date of the earliest activity that has a path to the target milestone
- Data date — Is the date at which the file has been actualised, this tells you “when” the forecast has been made
- Project name — The name of the project for which this report has been made.
- Milestones & Activity collation — Tells how many milestones and how many activities nPlan has read from the entire file that you have uploaded
- Milestones & Activities overview — States a high-level summary of the state of the tasks in the programme. In particular you can see how many of the tasks in the file are connected to the target milestone that you have selected, how many are disconnected, not started, in progress and completed.
- Requires attention — Contains selected integrity issues that are likely to affect the confidence on the correctness of the forecast. We recommend that you address these issues before continuing with taking action or else make it clear to your risk teams that there are high-priority integrity issues that could have an effect on the results. You can find more details about these in the following sections.
- Activity and milestone assessment — Contains an assessment of integrity that relates specifically to how activities and milestones have been recorded in the schedule file.
- Activities with 0 duration — Reconsider whether these should be deleted or if they are actually milestones
- Long durations — By default we consider an activity to be long if its duration longer than 60 calendar days or if its duration is longer than 10% of the entire project duration. Consider adding detail to your schedule by splitting this activity.
- Activities with only “to-finish” predecessors — These are usually free to move in the forecast because the start date is not linked to any predecessor. We recommend including at least one “to-start” link so that their uncertainty is included in the forecast.
- Activities with only “from-start” successors — The uncertainty of activities with only “from-start” successors has no effect on the forecast because their end date is not connected to any activity. We recommend including at least one “from-finish” connection for every activity.
- Deleted activities/milestones — We identify activities that have been marked for deletion within the schedule. We ignore these in the forecast, but we recommend removing them from the schedule file to avoid confusion about the scope of the project.
- Allowance activities — We identify activities that are meant to represent waiting periods, float or risk allowances. We dis-recommend the use of float for programmes on which you are forecasting, but if you must use float, please include it in the relevant field of your scheduling software.
- Proportion of activities that are resource loaded — The proportion of activities in the project that contain data about the resources used to scope the activity. DCMA-14 suggests that most activities are assigned the resources that were used to calculate their duration, you can use this field as a sanity check.
- Tasks with no relationships — Floating activities should be verified and connected in the adequate place in the programme.
- Tasks with successors and no predecessors — Ideally, only start milestones or activities should have no predecessors. We recommend reviewing this list and verifying that all activities are properly connected.
- Tasks with no successors and with predecessors — Ideally, only end milestones or activities should have no successors. We recommend reviewing this list and verifying that all activities are properly connected.
- Link assessment — We do a thorough assessment on the links that connect every activity in the schedule. Most of the links should have no float or lags in them and they should be finish-to-start in order to ensure a high-confidence forecast.
- Float and lags — Forcing float and lag in relationships between activities makes it hard to produce a high confidence forecast. This is because it is often ambiguous how this float or lag should be handled when applying uncertainties to the durations of the activities. We generally recommend that a schedule has as few as possible floats and lags. Note that sometimes these may be caused by constrained activities, see the next section for more details on constraints. We will provide warnings for the following float and lag issues:
- Negative float
- Negative lag
- Positive lag
- High float
- Cyclical logical relationships — If there is a loop in the logic of the programme, our forecasting engine will not be able to produce a forecast. This issue needs to be resolved before you can see any results.
- Start-to-finish links — These types of links typically represent an error in logic. We recommend to change all these relationships to a finish-to-start, start-to-start or finish-to-finish as required.
- Links with different calendars — We will flag when two linked activities have different calendars. If this number seems too high, we recommend to double-check whether mistakes were made when assigning calendars. Erroneous calendars can dramatically affect the results of the forecasting engine. So we recommend reviewing these thoroughly.
- Duplicate links — This indicates that there are pairs of activities in the programme that contain more than one link. Whereas there are valid cases for this, we recommend reviewing these and deciding whether the double-links are required.
- Link summary — We will present a pie chart showing the distribution of link types in the programme. DCMA-14 recommends that 90% of the links should be finish-to-start.
- Float and lags — Forcing float and lag in relationships between activities makes it hard to produce a high confidence forecast. This is because it is often ambiguous how this float or lag should be handled when applying uncertainties to the durations of the activities. We generally recommend that a schedule has as few as possible floats and lags. Note that sometimes these may be caused by constrained activities, see the next section for more details on constraints. We will provide warnings for the following float and lag issues:
- Constraints — A constraint in the schedule logic restricts our ability to apply uncertainties. Whereas we recognise that there are valid reasons to apply constraints to the schedule (e.g. due to contractual obligations) we recommend to remove all constraints when forecasting in order to get a transparent and realistic forecast. We will divide constraints into two:
- Hard constraints — These restrict the activity or milestone to a single point in time (i.e. must start/finish on and start/finish on). Results for activities or milestones which have hard constraints on them are invalid. So these should either be removed or changed to a softer constraint.
- Soft constraints — These that restrict the activity or milestone only on one side (i.e. start/finsh on or after/before). Results are still valid on these constraints as uncertainties can be applied up to the point where the constraint applies. However, we don’t recommend using these constraints as you would not be observing the full range of possible outcomes. In particular, if you are interested in potential threats to your project, we recommend removing all “start/finish on or before” constraints. If you are interested in identifying opportunities we recommend removing all “start/finish on or after” constraints.
- Project calendars — Calendars can have a large effect on the outcomes of the forecast engine. Factfile provides a list of all the calendars that we have found in the file and their main characteristics. We recommend that you take a look and investigate if something doesn’t match your expectations of which calendars are available in your file or how many activities are affected by that calendar
- Inconsistent completion — This section highlights activities that may indicate erroneous recording of actuals. Because we don’t forecast on actualised data, making sure that activities are not actualised when they shouldn’t ensures that your forecast is reliable. We will highlight activities that have been actualised after the data date and activities that have been actualised but whose predecessor has not.