You can use the AI configuration application to AI to find work orders that are similar
to another work order. For example, if applied to work orders, you can find work orders that are
similar to another work order. The similar work orders feature uses the similarity model
template.
Before you begin
If you are using Maximo® AI
Service on-premises, review the entire process
for enabling AI features and complete all prerequisites, including deploying Maximo AI
Service. For more information, see Maximo AI Service and AI features in Maximo Manage.
If you are using Maximo AI
Service SaaS, ensure that you have specified
values for the Maximo Manage system properties. For more information,
review your welcome letter.
By default, the similarity feature is enabled for all work orders that are not cancelled or
closed, have a work type specified, and the work type is not PM. You can change what work orders are
included by editing the query definitions for the MXAPIWODETAIL object structure. For more
information, see Customizing integration components for locating similar work orders.
You can activate and test the AI configuration for locating similar work orders and then change
the filter later.
About this task
The WOSIMILARITY AI configuration is available by default. The configuration uses preconfigured
Maximo Manage integration components, such as an object
structure and invocation channels. You do not have to configure these components to set up the AI
configuration. Later, if you want to fine-tune how data is transferred or managed, you might alter
some aspects of the components at your own digression. For more information about the components,
see Integration components for
locating similar work orders.
The AI configuration for finding similar work orders does not require training but does require
some automated configuration processes after activation and prior to inferencing.
Procedure
- In
Maximo Manage, in the AI configuration
application, in the table, click the WOSIMILARITY AI
configuration.
If you delete the default configuration and must add another configuration
for the same feature, ensure that you use the similarity model template and that you select the
latest version of the template.
- In the Edit AI configuration dialog, review the default settings.
- Click
.
- Optional: In the Related
Records field, specify a WHERE clause to limit what work orders can be reviewed for
similarity.
For example, you can limit the search to work orders that are for a certain
asset or of a certain type. If a work order does not meet the criteria of the WHERE clause, you can
trigger the similarity feature from that work order, but no similar work orders are
returned.
- In the Template version field, ensure that the latest version
is selected. To view all versions, click the Lookup icon.
- Optional: In the Additional details for AI explained section, provide any
information that is specific to your organization and relevant to your end users to help them
understand the model and its output.
An AI icon is located
alongside your model's output. Your users can click the AI icon and then
access any information that you specify in this section alongside other general model information
that's provided by IBM. You can complete the AI configuration process, view the
AI icon and its content in context, and then edit this section later as
needed.
- Click Save.
- Set up arguments.
- Click .
- For each of the arguments, review any default value and change as needed. Specify
values for the date_field and features arguments. The
following table describes all of the arguments.
Table 1. Arguments for similarity model
template
| Argument key |
Description |
Type |
Default value |
| date_field |
A date field in work orders that is used to determine similarity. If you change the value for
this parameter, you must also change the query templates for similarity in the MXAPIWODETAIL object
structure. For more information, see Integration components for locating similar work orders. |
string |
changedate |
| features |
Fields to compare to determine how similar work orders are. If you change the value for this
parameter, you must also change the query templates for similarity in the MXAPIWODETAIL object
structure. For more information, see Integration components for locating similar work orders. |
string |
description, description_longdescription |
| history_days |
The maximum age in days of work orders that can be included for inferencing. This value is
calculated from the base work order's date field value. |
integer |
30 |
| max_similarity_records |
The maximum number of similar work orders that can be returned. Note: In the System properties
application, you can configure the mxe.ai.comparetolimit property to determine the
maximum number of work orders that can be reviewed for similarity.
|
integer |
10 |
| retention_days |
The maximum age in days from the current date for work orders that can be included for
configuration. Work orders can be at most 1000 days old, but the higher this value, the longer it
can take to complete inferencing and render results. This value also determines the age of work
orders that can be similar. For example, a work order with a change date of today can return similar
work orders that have a change date of up to 90 days ago. |
integer |
90 |
| similarity _threshold |
The minimum similarity score that work orders must have to be included. |
integer |
70 |
- Click Save.
- Click
.
Activating the AI configuration indicates that the AI configuration is prepared and the model is
ready to be processed by the cron tasks. Activation also automatically runs configuration and
inferencing.
- Click .
What to do next
Inferencing is controlled on a crontask schedule. The AIINFJOB crontask initiates inferencing for
all eligible AI configurations. By default, the cron task runs every hour. If inferencing does not
start promptly after the model is activated, you can edit the cron task schedule:
- In
Maximo Manage, in the Cron Task Setup application, open
the AIINFJOB cron task.
- In the Cron Task Instances table, for the WOAI instance, in the Schedule
field, change the value.
For more information about the cron task, see Cron tasks for training and inferencing.
If you want to reconfigure the model on new data in the same filters, such as changed query
definitions, or you want to edit the arguments, make any changes and then in the AI configuration,
click .
If you need to change other configuration settings, you must first deactivate the configuration.
In the AI configuration, click
, edit the
configuration and then activate the configuration again. Activating the similarity AI configuration
automatically starts reconfiguration and inferencing.
Access to the similarity feature is controlled through a signature option for the associated
object structure. By default, the MXAPIWODETAIL object structure is used. In the Security Groups
application, select the security group to which you want to grant access. On the Object
Structures tab, in the Object Structures table, select the object structure that you
associated with the similarity AI configuration. In the Options for table, for the View
similar work order option, select the Grant Access checkbox.
For more information about using this feature to locate similar work orders, see Viewing similar work orders.