Import AI-Driven Suggestions From Schedule History
Use Case
A team member is planning out the design work for the piping portion of the schedule. He draws from past projects to build out his high-level structure to save time and make sure his plans are accurate.
This section covers how to draw from InEight Schedule’s knowledge base of past work to build out your schedule, based on AI-driven schedule suggestions.
Knowledge Base
The Knowledge Base is the repository where master data for your schedules is stored. This includes historical or template CPM schedules that can be leveraged for creating new plans and benchmarking your work.
You can publish all or part of an existing schedule to the Knowledge Base by selecting the Actions ellipses at the level you’d like to publish. For example, to save the entire schedule, you can select Actions at the highest planning level (project level).
The schedule scope will need to be verified by the Schedule Administrator before it can be used for schedule suggestions.
Inference Engine
This stored schedule data feeds InEight Schedule’s Inference Engine. When building out a new WBS item or activity in a schedule, this Inference Engine uses AI or machine learning to draw from your stored data and make suggestions based on how well the data matches your planning package or activity.
The Inference Engine uses certain criteria to determine what data is most relevant to suggest, such as description and phase. Within the Knowledge Base, the influence of each of these criteria can be increased or decreased to tune or calibrate how the Inference Engine makes suggestions, under the Machine Learning tab.
Knowledge Tags
Knowledge tags are labels you can set up to further define and contextualize your data. The Inference Engine will also take these tags into consideration when determining schedule suggestions.
Knowledge tags are set up and stored in the Knowledge Base where they are organized by code or user defined field (UDF) at the project, activity, or resource level.
For example, you may use a Project code called Location. You could edit this code to define the following values:
When building a new schedule for a North American project, the Inference Engine will look for projects labeled with the North America tag value when making suggestions.
You can also use the North America tag to filter down to just North American projects in your view.
Access the following links to learn more:
Review Schedule Suggestions
When building your conceptual schedule, you can leverage the schedule suggestions provided by InEight Schedule’s Inference Engine.
Each schedule suggestion is a knowledge subnet, or sequence of planning packages and/or activities.
Let’s look at an example of how this works.
Step 1
You are planning out Piping design work and want to review schedule suggestions based on similar work done in the past.
In this case, you select the IFC > Piping scope level.
From the Iris side panel, you expand Smart Planning to see the schedule suggestions.
Select and Review Knowledge Subnet
Once you’ve decided which suggestion to import, you can select and review it prior to importing.
Step 1
After review, you determine the best suggestion to go with and select its Import Knowledge Subnet icon.
Preview and Import Knowledge Subnet
The system allows you to preview what the subnet will look like merged into the schedule prior to import.
Step 2
The schedule displays in Preview Mode, showing the new subnet merged into the schedule.
In this case you only need the first three Spool activities, so you remove the other three prior to importing.
Step 3
Select No to cancel the import.
Select Yes to proceed with the import of the subnet.
In this example, you select Yes.
Access the following links to learn more about building out your schedule using AI-driven suggestions:
Back to Scheduling and SIP main page