====Control Example: Time Spent Out Of Control====

In this example, we see if our machine is maintaining the oven temperatures we are targeting.

===Available Information===

^Asset^Property^Desc^
|Finish Oven Zone 1|Temperature - Oven|The actual temperature of our oven|
|Finish Oven Zone 1|Temperature - Target|The expected temperature of our oven|

===Initial Layout===

{{control_start.png}}

To begin with, we have one [[control_output|Control Output]] node, one [[control_control|Control]] node and two [[ardi_point|ARDI Point]] nodes.

===Step 1: Determine Control===

  * Click on the top **Measurement** node, where it says //none//. \\
  * From the pop-up list, choose the //actual value// that is being controlled.
  * Click on the lower **Target** node where it says //none//.
  * From the pop-up list, choose the //target// that you're trying to reach.

{{control_basicpng.png}}

===Step 2: Tune===

{{control_basic_pretuning.png}}

When we save this and look at the report, we can see that it's almost **always** out of control.

This might be true, or it might mean that our tolerances could be a bit too strict. 

If we go in and update our **span** to be within //5 degrees// of the target instead of //1//, we get the results below...

{{control_tolerance_update.png}}

This is //better// but still not great. Either our oven is having real issues, or we need to refine our logic further.

We can see a lot of noise - going in and out of control very frequently. But there's the opportunity to remove some of these small periods of control loss with a **timer**.

{{control_basic_withtimer.png}}

This updated logic tells the system that the measurement has to be uncontrolled for four seconds before it's considered being genuinely //out of control//.

===The Results===

This results in a chart that is **far** less noisy, although still showing some issues. 

{{control_basic_final.png}}

===Deployment Options===

Control charts have one of the largest lists of [[control_deployment_options|deployment options]].