The reason for MD's questions is that a timeout usually occurs because the time allocated for the task has been exceeded. By default, MOST tasks are given 30 minutes to finish. If the time to image exceeds the time allocated, it will "time-out" in the console, while still recording a success. The failure appears on the Jobs and Tasks main-screen, but a success will appear if you drill-down into the task. Weird, but there it is.
The problem with this logic is that the default timeout for a deploy image task is 960 minutes (non-editible, found if you right-click the task and select "view as XML").
Which then begs a few other questions. 1) you say they use backup images - are they distributing these to more than one workstation? They should not. 2) are they using a deploy image task, or a custom script task to deploy the image - such as calling RDeploy yourself, or even ImageX?
Note that the default timeout for a script task is 30 minutes, but it's editible, and found under the Advanced tab. IF you are using a script, up the timeout to 60 or 90 minutes - well in excess of what it actually takes, and the problem will be gone.