Using OpenSched
This is supposed to demonstrate how you could use opensched during
project planning and project tracking.
- Create a work breakdown structure, generating a list of
activities and a list of tasks to be performed for each activity.
- For each task, estimate the number of days effort, for some
'standard' person (e.g., the average if all people in your
organisation, the best, the worst, yourself). Be clear what a day is!
To OpenSched, days are elapsed time. So if, for example, you think in
terms of hours and have a task that could take your standard person 7
hours, he/she works 7 hour days but only speeds 50% of them being
productive, then the task is 2 days effort not 1 day.
- Identify task dependencies.
- At this point, you are ready to create a .sched file. Enter all
the tasks. Create dummy resources for each task unless you know who
will be doing them. Enter the task dependencies.
- Then run opensched. This gives you your first schedule, which you
are obviously unlikely to meet unless you have lots of available
resources.
- Add some more constraints, such as:
- Public holidays
- Adding real resources, remembering to include any appropriate
vacations and setting efficiency. Then assign tasks to them.
- Identify tasks that could be sped up by having more than 1 person
working on them simultaneously. Either replace the task with multiple
smaller tasks, or replace the assigned dummy resource as with one representing
a group of people and increase the efficiency to reflect this (e.g.,
setting the efficiency to 2.5 times the average if you assign 3
resources). If the latter, make the resource name reflect this
group.
- Reiterate through the above step, perhaps also changing the
resource assigments (HOW DOES OPENSCHED CHOOSE RESOURCES) until all
dummy resources have been removed, all dependencies have been
identified, and the task duration estimates are correct.
- Baseline the schedule (HOW?).