In the CMDB every CI has a Instance ID that is a GUID that is unique in the database (and should be unique everywhere). This value is auto generated on creation unless the once creating it specifies one (I think this rare) and it is required to be unique.
I do not know if asset uses this value for the asset id but it would make sense.
The instance ID is not exposed to the Asset Forms. It is visible only to the BMC_Core backend forms. This number is rather long ( OI-8EAA41A18D384737A6A56856DAC94433). You could create workflow which pulled this value from BMC_ComputerSystem and did a push to the corresponding record in AST:ComputerSystem.
The downside of this is that it would be very difficult to see the difference between two such numbers.
The InstanceId field IS available on the AST:XXXXXX forms as they are just self joins of the corresponding BMC_CORE:BMC_XXXXX forms, and InstanceId is on every one of those class forms.The problem is that it is not in the views by default. So no need to create workflow to push values from one form to another, BUT you will need to add the field to the views required, AND then push that field value into Asset ID.
The bigger question which Curtis mentions above is the value of using InstanceID. It's not a very friendly entity.You could use AssetID (aka CI ID) with the CI name and the status e.g.in your case, perhaps servername:status, AU101:Deployed, AU101:Disposed etc. might work??
I guess it depends on specifically why you need that unique identifier, where it will be used, and whether it's something that you would like to be able to genertae automatically in soem cases etc.
Thanks guys but I guess what I'm looking for is not only a recognised and consistent convention for CI ID but also a mechanism for populating it, automatically if possible. e.g. if I start by saying my first CI is CI0001 and I continue on this path what is to stop me duplicating an entry or skipping one out of sequence? Any ideas? What is a typical convention for CI ID?