There's a discussion of using Subversion (svn) for synchronized backups of this sort at:
http://stackoverflow.com/questions/61888/using-subversion-for-general-purpose-backup
Also, one of the entries in that discussion mentions bup, a git-based backup tool...
https://github.com/bup/bup
You can schedule periodic tasks on the different network boxes to run commits to a repository on your machine (or elsewhere). That repository should be defined on the directory tree where you want your local copy to reside on your box. Requires that you:
- install svn (or git) as a service on your box
- install as a client on the other boxes
- create batch task to run on each network box to update from and then commit to your repository
I don't use git, and haven't automated the use of svn for the purpose. However, I've used svn for years both professionally and privately and love it. I use svn on my home network to keep my code synchronized across my three development boxes, i.e., workstation, laptop, and netbook. I also keep several directories of personal information in my svn repository both for convenience of cross-box access, version control (like the original intent of a version control system), and backup safety.
Note - I have a separate server on my home network for this...just something you might consider. My svn (and print) server is an old, 1.2GHz Pentium with only 512MB of RAM and about 120GB of storage. No keyboard, mouse, or screen - remote administration only. It's a very cheap solution that ensures the repository is always available when the clients want to update/commit. If you're keeping your repository on your own machine, your box has to be on and connected to the network in order for the other boxes to connect to the repository. If you install svn as a service on another box on the network, you can simply update you own working copies of the files on your box as a client whenever it suits you to be connected, and all other clients can update and commit on any schedule, presuming you leave the server always-on.