Thoughts on tracking large engineering projects

 Here's a short summary of my thoughts on efficient tracking of large projects that I have used with some success and gradually refined over time. 

1. Only track milestones.
2. Minimize the number of tracked milestones.
3. Make the dependencies between milestones explicit.
   This requires minimal tooling.
4. Express upstream dependencies only.
   A milestone owner must document what they depend on, not where their deliverable will be used.
5. Do not track effort.
6. Do not track offsets from milestones.
7. If a milestone’s date is in the past, assume it to be done.
8. Track changes in committed milestone dates. This is the weekly tracking that is needed.
9. Build tooling such that every team only updates their own milestones and the tooling alerts the dependent milestone owners of changes.

This can be all done with Excel and minimal IronPython. There is a place for rigorous tracking using MS Project for tasks and resources, but in my role as an engineering lead, I need to react to movement of large milestones and I've found the above methodology quite useful. 5 and 6 are bread and butter for program managers and 7 borders on heresy but for a lead engineer these guidelines do help reduce the management time spent on things that are on track.

What are your best practices?

Regards,

Kuntal.