Coordinating so many people, and so many antennas

With hundreds of people working around the clock for years with the goal of creating the largest observatory in the world by 2014, it can be a challenge to make and track progress in this astronomical marathon.  Sitting in the control room for 12 hours each day, I see the hustle and bustle of many people performing many tasks, from telescope operations to antenna engineering tasks, often with success but occasionally with some failures as well.  It's important that everyone remain safe (for example, if someone is moving an antenna at the testing site outside, no one should walk over there unaware), that jobs are performed efficiently (i.e. two people don't perform the same redundant task unknowingly), and finally that jobs and problems be prioritized.  In a sense, communication is the key. 

Photos from almaobservatory.org

How do hundreds of people, each with different areas of expertise, communicate effectively?  One way is by creating layers of organization, including the ALMA board, management, committees, and teams of staff.  For example, I'm a part of the "Joint ALMA Observatory Science Team", and I work in the "Commissioning and Science Verification" group.  We meet every weekday at 3 pm to discuss the status of the previous day's projects, and plan for the day/night ahead.  It's certainly easier to get everyone together for 30 minutes and explain everything once, rather than having many small discussions which need to be repeated throughout the day.  Once per week, we have a Science Team meeting, when all of the groups come together for more coordination. 

The ALMA board in one of their important meetings in Santiago.  This room can connect via telecon to a meeting room at the OSF.

When I say "together", I mean this rather loosely, because at any give time, some of the team is in Santiago at the "Santiago Central Office (SCO)", while others are at the ALMA "Observations Support Facility (OSF)".  We connect via telecon, and it feels like one meeting room merges right into the other, only separated by a projector screen.  

Another innovative way that we stay connected and coordinated is via an online commercial software product called "JIRA".  The name is actually based on the similar bug tracking program "Bugzilla", which became called by the Japanese name for Godzilla, pronounced something like "go-jira".  How JIRA works at ALMA is that any time there is a problem that needs to be addressed, someone makes a "ticket" for the issue, and assigns the issue to someone on the ALMA team.  As the issue is addressed, it's important to keep good documentation in the form of comments and attachments on the ticket's website, and when the issue is resolved, the ticket is closed.  The tickets are helpful for letting others within the group, and also in other groups, what are the most important issues for developing and maintaining the telescope system (hardware, software, and anything in between).

Jumping into the CSV project head first, I was assigned several tickets and was able to determine their status, following the comment trail left by people who worked on the same problems recently (some of whom were visitors, or work elsewhere now).  Attached to the tickets were reports, programs, and information on how to access relevant data.  When the tickets were re-assigned to me, I even got emails from others who had been involved, encouraging me to ask questions whenever something wasn't entirely clear.  In this way, projects (especially long-term projects) are less likely to fall off the radar completely, even with many people working on many different projects, because the goal is to ultimately "close" all of the tickets in preparation for the fully-operational ALMA. 

The OSF Control Room is a very collaborative environment, and a great place to ask for help or explanation of the ALMA system.

Today I made my first JIRA comment, updating others on what I had accomplished (or rather, where I found a bug) with my project.  I'll keep working to fix the bug, with the goal of "closing" the ticket as soon as I can, and moving on to another.  I sometimes stop to think about how remarkable it is that ALMA can progress, considering the scope of the project.  But, when I think of the project as being comprised of teams with many dedicated and capable people, each closing issues one JIRA ticket at a time, I think that anything is possible.