====Outage Report Example: Simple==== In this example, we'll look at the outages we've experienced on our Paint Line, and try to classify why some outages happened. ===Complexity=== Because this is more complex than some other metrics, this tutorial isn't quite as step-by-step as the previous examples. It's assumed that you know how to add a node and are familiar with some of the basic node types. If you're new to the visual analytic system, we suggest looking at the [[tutorials|tutorials]] for simpler modules, such as [[module_control|Control]] or [[module_pivot|Pivots]]. ===Available Information=== ^Asset^Property^Desc^ |Paint Line|Status - Painting|Has a value of **1** when painting, and **0** when stopped.| |Paint Line|Actual Speed|The speed the machine is running| |Paint Line|Thickness|The thickness of the product being produced| |Paint Line|Width|The width of the product being produced| ===Initial Layout=== {{oee_trinity.png}} The key nodes are the three on the far-left. We need to provide logic for how the system can tell that the asset is... * Available (should be able to run if requested), * Running (actually running), and * Required (needed to run) ===Step 1: Available, Running & Required=== The first step is to choose the rule we will use to tell if the machine is //available//. While the [[example_oee_basic|OEE example]] has more complex logic, this time we're looking at the system as a whole. It's **always** expected to be available and running. It's actually **is** running when the speed is more than 10 meters per minute. The logic for those rules looks like this... {{outage_basic_initial.png}} ===Step 2: Processing Events=== The Outage Reporting analytics have a special [[input nodes|input node]], called [[outage_eventcompare|Event Compare]]. This node is run //after// all of the categorisation has taken place. Then for each different outage time-period that was detected, the Event Compare node is run. This lets you compare the values at the **start** and at the **end** of your event to see if you can get any other insight into what type of outage it was. In our example, we can compare the value of **Product Thickness** at the //start// of our event with the thickness at the //end// of the event - and if it's changed, we can mark the event as a **Thickness Change**. {{outage_basic_refine.png}} The Event Compare node has three outputs - the //start// offset, the //end// offset, and the //duration// of the event. The offset amount is in samples relative to the start of the query time, and can be used with the [[ardi_pastpoint|ARDI PastPoint]] node to read a property value from a point in the past.