This document outlines an alternative approach to including a custom/external JDBC driver. The method outlined in the installation guide is to manipulate the aveksa.ear file, which is then deployed on top of wildfly. That method is best when running on a wildfly cluster, or if you're already modifying the ear.
This method, on the other hand, simply loads the JDBC driver as a module within wildfly, which is then useable by all wildfly applications (including your collectors). It works best when a simple approach is desired.
In the document, the jTDS driver is used as an example.