AirVision’s optional patented Automatic Data Validation Processor (ADVP) allows an unlimited number of logical rules to be applied to data, resulting in data that is flagged, graded, annotated, coded, and/or emailed to designated users.
Click here to see a video demonstration of the ADVP.Each ADVP rule consists of two main components: a trigger and a resultant action. If the conditions of the trigger are met for a given data point, the action is taken; if not, nothing happens, and the next rule is evaluated. A trigger consist of a number of conditions linked by logical combination (e.g., AND/OR).
The following is an example of a data rule:
If Ozone is greater than 50 ppb and the ambient temperature is less than 72 degrees F, then mark the suspect data.
The
trigger is composed of the following two conditions:
If Ozone is greater than 50 ppb (Condition #1)|
AND
the ambient temperature is less than 72 degrees F (Condition #2)
The
action is:
mark the data suspect data
>>Note: multiple AND/OR links are evaluated in sequential order. For example, a four-condition trigger such as:
Condition 1: 1 > 2 link AND
Condition 2: 2 > 3 link: OR
Condition 3: 3 > 4 link: AND
Condition 4:
Is evaluated as follows:
(((Condition1 AND Condition 2) OR Condition 3) AND Condition 4)
in other words, having Condition 3 and 4 as true would cause the action to be taken.
Each trigger condition uses a Parameter Template, and optionally can be associated with a particular site. If the site field is left blank, the rule is applied to all sites with parameters that match the templates.
The trigger condition can use certain characteristics and make certain comparisons:
| Characteristic | Comparisons Allowed | Compared To |
| Average Value | Greater Than Less Than Equals Differs Greater Than Differs Less Than | Another Parameter (Template), same site or specified site.
Historical Average (fixed historical averages- see end of chapter)
Historical Composite (a dynamic average of previous years of the same date/hour, similar to the Historical Composite used in the Average Data Editor. |
| Flag | Contains Doesn’t Contain | List of logger/system flags |
| Annotation | Contains Doesn’t Contain | User-defined text |
| Variance (change from a previous average value) | Greater Than Less Than Equal To | User-defined percentage |
Example (Figure 6-1): To do a simple wind vane stuck rule, you could point at the parameter template for WDR and set the trigger condition to be Variance, Less Than, 5%. The rule can be enhanced by using the
Persist Rule check box under
Rule Details to require the condition to hold for a certain number of averages.
To create the rule, click the
New Rule button on the ribbon, and we give the rule a name and enable it. Designate the average interval, and hit the
Save button to add the first condition by clicking the
Add Condition button at the bottom.
Fill in the first condition and select
Variance for the
Characteristic, set the
Relationship to
Less Than, and designate the
Parameter Template to apply. Finally, set the limit for the comparison to 5%.
Figure 6-1 Stuck Wind Direction Program in ADVP Editor
The rule can be further improved by adding another condition that the wind direction be below a value, for example, 10 mph, to avoid cases of steady winds accidentally triggering the rule. Set the
Relation to Next Condition at the bottom to
AND, and use
Add Condition to create , filling in the values and designated WSP for the comparison parameter template. Since you are using parameter templates, this one rule can be applied to all sites where WSP and WDR labeled parameter exist (Figure 6-2).
Figure 6-2 Further improving the ADVP setup
Actions can consist of any or all of the previous steps, and they can be applied to any parameter (template), not necessarily the one that triggered the rule (Figure 6-3):
- Assign a Fixed Value
- Apply a Flag
- Clear a Flag
- Apply a Null Cole
- Clear a Null Code
- Add Annotation Text
- Assign a data Grade (1 .. . 10)
- Send an Email via Subscriptions Notifications
The action can be applied to any parameter at the same site, or a specific site can be designated (most likely approach). In the previous example, you could mark data invalid and set the null code AW for Wildlife Damage. You could optionally email QA, in case the sensor failed.
Figure 6-3 Setting Null Code and emailing QA as a result of ADVP rule