After you configure the Parts Identifier add-on in Maximo Application Suite, you must connect Maximo Manage to the Humai Cloud service to send data to the service. In addition, you must send images of the Maximo Manage parts and items to the Humai Cloud Service to service to train an AI model to identify images.
Connecting Maximo Manage to the Humai Cloud service
You can connect Maximo Manage to the Humai Cloud service by using multiple communication modes that include web services, Java™ Message Service (JMS) messaging,
Apache Kafka messaging, and hypertext transfer protocol (HTTP). This procedure describes how to connect Maximo Manage to the Humai Cloud service by using Apache Kafka. For more information about connecting Maximo Manage to external applications, see
Integrating data with external applications.
1.Record the following configuration details for your Strimzi cluster. You need this information to create a message provider for Kafka in Maximo Manage:
- BOOTSTRAPSERVERS
- SASL_MECHANISM
- SECURITY_PROTOCOL
- USERNAME
- PASSWORD
2.Open a browser window and download the Kafka cluster-ca-v0 certificate from your Strimzi cluster. You need to add this certificate to Maximo Manage in step 3.
3.Complete the following steps to add the Kafka CA certificate to Maximo Manage in Maximo Application Suite.
a.Open the Kafka CA certificate in an editor and copy the contents.
b.From the Maximo Application Suite catalog, on the Applications tab, select the Manage tile.
c.Click Actions>Update Configuration.
d.Click the edit icon for Import Certificates.
e.Click Add Another.
f.In the Certificate Contents field, paste the contents of the certificate
g.Click Confirm.
h.Click Activate.
4.In Maximo Manage, complete the following steps:
a.In the External Systems application, add a message provider for Kafka. For the properties, specify the configuration values for your Strimzi cluster that you recorded in step 1.
b.In the External Systems application, add a queue. In the Message Provider field, specify the name of the Message Provider that you created for Kafka in the previous step, for example, STRIMZI. In the Queue JNDI Name field, select the name of the Kafka topic.
c.In the End Points application, edit the preconfigured Partium endpoint. Specify the URL, username, and password that were provided by Humai to connect to their service. Click Test to verify the configuration.
d.In External Systems application, edit the pre-configured Partium external system. In the Outbound Sequential Queue field, specify the name of the Kafka topic that you configured in the external systems queue. Enable the Partium external system.
e.In the Cron Tasks application, create a cron task for Kafka.
Training an AI model to identify images
In the training phase, the Humai Cloud service learns to identify images of parts that are uploaded to the service in JPEG or PNG format. For training, you must also provide a mapping file that associates each image file that you upload with a Maximo Manage part. A mapping file must have one of the following formats:
|
File Type
|
File Format
|
|
CSV
|
Column 1 contains the unique part ID for the part. Column 2 contains the URL or relative path to the image file.
|
|
JSON
|
The key contains the unique part ID for the part and the value contains an array of URLs or relative paths to the image files.
|
Where, depending on the object type, the unique part ID is a pair of Maximo Manage attributes that are separated by a colon:
|
Object Type
|
Attributes |
|
Item
|
<ITEMSETID>:<ITEMNUM>
|
|
Asset
|
<SITEID>:<ASSETNUM>
|
For information about how to upload the images and the mapping file to train a model, contact Humai by using the support@partium.io email address.
[{"Type":"MASTER","Line of Business":{"code":"LOB59","label":"Sustainability Software"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Product":{"code":"SSRHPA","label":"IBM Maximo Application Suite"},"ARM Category":[{"code":"a8m0z000000cwZDAAY","label":"Maximo Application Suite"}],"ARM Case Number":"","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"8.4.0"}]