When considering internal process auditing metrics I like to think about performance.
This means going war far beyond how many audits are performed, how many findings are recorded, and even how quickly the fixes are inplace. These things do not tell me anything about how the audit process is helping the organization to achieve its goals.
Let us draw a distinction between the kind of auditing we are referring to. I am referring to system audits - audits of processes, as a registrar would do except in rather more depth. What kind of auditing are you referring to?
If there is a new process or work group, having the cell of people audit themselves frequently (once per shift) can be worthwhile if the audit is a limited check of a sample or performance to a procedure, and is overseen by an experienced authority.
A weekly supervisor audit can be useful if the standards of expectations are well defined. The management can also do these as long as no contradictions between the "auditors" result. The he said/she said effect can be very damaging. This MBWA
method has been around for quite awhile. It should not be called an audit, as the term audit
implies there are stringent requirements and punitive consequences.
I have heard of stop light signal charts - they were well received when used constructively.
has a paper that might help you.