Skip to content

Scheduling Optimization and Machine Learning

Scheduling Optimization and Machine Learning is a built in capability for IFS Cloud to interact with IFS Planning and Scheduling Optimization (PSO). IFS PSO manages the process of assigning activities (jobs, tasks etc.) to resources (people, machines etc.) in the most efficient way possible, according to a defined set of constraints. At the core of IFS PSO is the Dynamic Scheduling Engine (DSE). The DSE is a generic optimization engine that finds the best solution to the problem that is given to it, whilst obeying the defined constraints. The DSE has been used to optimize geographic field workforce routes, manufacturing, timetabling and aircraft scheduling.

In addition to the core scheduling capability, IFS PSO also supports a number of additional functions:

Appointment Booking is the process of selecting a suitable time slot for an activity based on an existing schedule. The software can suggest which slots fit best into the existing schedule and so would be more preferable.

Scheduling Data Management is where data can be set up and referenced that is specific to scheduling and so may readily available in IFS Cloud. This includes data such as activity types, sla types, regions, skills, parts, appointment templates etc.

Resource Planning is the process of planning when resources will be working and so creating a rota. It is possible to automatically create a rota based on a set of requirements to govern what type of resources are needed and when, and also rules to prevent resources working too much. It is also possible to ask the DSE to create a rota automatically.

Scenario Planning allows customers to see how changes to their activities and/or resources will affect their ability to meet their performance targets. In addition the software can suggest what changes can be made to the current workforce in order to meet these targets.

Machine Learning allows defining and training machine learning models using data from IFS Cloud and running inferences on trained models directly from IFS Cloud business logic.

Further details regarding PSO can be found in the PSO release documentation.

Interaction between IFS Cloud and IFS PSO

The integration between IFS Cloud and IFS PSO is initially configured in Scheduling Optimization and Machine Learning Configuration page.

The data interaction and mappings are developed by creating a Scheduling Model for each usage area using the IFS Developer Studio tool. Read more about developing Scheduling Optimization and Machine Learning models here.

The Scheduling Model consists of the following data interaction sections:

PL/SQL Code and Triggers are generated from the Scheduling Model and deployed to the database.

To use the Scheduling Model a dataset will need to be configured and enabled in Scheduling Optimization Datasets page. This will generate a Database Task that sends the scheduling data from IFS Cloud to IFS PSO as shown below:

A Machine Learning Model is currently defined as a Scheduling Model. This may change in future releases. To create a Machine Learning Model follow the same process as for creating a Scheduling Model.

The Machine Learning Model consists of the following data interaction sections:

  • Machine Learning Data - Used for machine learning data integrations.
  • System Data - Used for system data integrations.

To use the Machine Learning Model a dataset will need to be configured and enabled in Machine Learning Models page.