Help brainstorming a solution

Scenario:

1. We do all our work in Java with cook as our DMT
2. We need to add Android development to our list of supported
architectures (we are handling this outside of the aeca(4) framework
and want to keep it that way)
3. We keep our proprietary  code on server A and code for a particular
client on B (there are other companies that are also on the team and
we do not want to reveal the  proprietary source code)
4. Due to item 2 the actual compile/install of an android app has to
happen on my local machine (we use FreeBSD but the team doing the
front end code for the android is on Win 7)
5. The servers and local machine(s) are separated by an ocean (and no
local machine is in the same physical location as any of the servers)
6. We want to keep this all under aegis in some form or an other

Here is the partial solution we have worked out for keeping the
servers in sync but what we are missing is the local machine stuff:

1. On A the default target of all builds is [project].jar
2. When a new change set of the project is complete and ready for use
on B we manually FTP the jar and treat it as a change/project file

What we need to do on the local machines (there is type A which is
internal ones to our company and B is for other team members):

A & B machines:

1. Make it so local change sets can be made and "pushed" onto server B
(review/integration is still manual there) [this is likely solvabole
with automating aedist in some way]
2. Make it so all local baselines are updated on integration of a change set

A machines only:

2. Make it so the A&B cases can be used for an arbitary project on server A

Other ideas?