Clone gets created with a status of 'on hold', and the change request shows the clone as failing before the scheduled date of the clone has even passed.
The change request will show the clone failed with error "target Instance is not valid" even though target instance is valid
The clone did not start prematurely but was cancelled due to a known problem PRB1383681 "Clone : Clone request in the instance_clone_schedule can go on HOLD with error "target Instance is not valid" even though target instance is valid".
This problem manifests when there is maintenance on datacenter in relation to the target's instance CI. The CI goes into "Repair in progress" state and this tricks the source instance into thinking that the target instance is not in a valid state and cancels the request. The clone appears as on "hold" in the UI and the associated change gets cancelled.
The workaround for this issue, is to request a new clone, which should not have the same issue again.