Follow

About work order insights

Before you begin

This feature is available to customers on the Professional and Enterprise tiers. To learn more, see How to sign up for work order insights.

Work order insights is an applied AI solution that automates the process of reviewing work orders to identify problems and abnormalities that could impact your operations. The tool predicts asset failure and reduces downtime, while helping you optimize the maintenance process to free up team resources. It sends two reports (one for open work orders and one for closed) directly to your email inbox, meaning that you don't have to spend hours analyzing maintenance data to uncover these issues.

To find abnormalities, the tool first groups or "clusters" work orders based on their description. This lets us compare work orders that represent similar work, so that we're not comparing (for example) replacing a filter with fixing an engine. This clustering process happens across sites, and is language agnostic, which means that we can cluster similar work even if it happened at different sites or was recorded in different languages.

Next, we compare each work order to the other work orders in its cluster to see how different, or anomalous, it is. Depending on which abnormalities we find, we'll assign a different risk type to the work order. For example, if we find that the work order was delayed longer than other work orders in its cluster, we would indicate that it represents an abnormal delay. To learn more about the risks we identify (and how they're assigned), see Risks identified in work order insights.

Lastly, we assign a risk score (from 0-1000) to each work order that represents how much it differs from the others in its cluster and the impact it can have on your operations.

To learn more about work order insights, see the following articles:

And be sure to check out our work order assistant, which helps you determine what steps to take after receiving the work order insights report:

The Work Order Assistant home page
Was this article helpful?
0 out of 0 found this helpful

Comments