This may be a simple question but I cannot get a set of activities to start on the same date. When I assign the relationship and put a Start to Start relationship the only option is Start on or before. I want all the activities to start on the same date. So the latest activity available will drive all of the start dates.
Am I missing something?
Thanks
For this purpose using SS links is not appropriate. The SS option means you got to link all predecessors to the first activity in the train. It not only will hide true logic but also create problems with resource leveling algorithms by not allowing any one activity to start before the other if availability of resources is scarce. It will force the resource leveling to follow the pre determined sequence in your SS links in a less efficient allocation.
Still you have several options. Say you have 3 activities named A, B and C and:
Logic predecessors of A are A1, A2, A3, A4 ... A8
Logic predecessors of B are A3,B1, B2. B3... B9
Logic predecessors of C are B3, C1, C2, D, E
If you want all activities to be scheduled to start at the same time you need to make all predecessors common to A, B and C. You have the following options.
Option 1: Link A, B and C directly to all predecessors. This I do not like as you loose perspective on why such links.
Option 2: Create a single milestone, link the milestone to all predecessors and to A, B and C as successors. Again this I do not like, the perspective is also lost but will save you a lot of links.
Option 3: Create 3 "predecessors" milestones one for Predecessors to A, another for Predecessors to B and another for Predecessors to C. Then create another milestone and link each "predecessors" milestone to a single milestone to mean Predecessors to A, B and C. This will keep logic transparent and in case you need to make changes it will be easier. I do not like this as it requires many additional activities just for this purpose, milestone activities that must be updated with actual dates when appropriate. On a simple job is easy to do but not always.
Option 4: If you have software capable of linking Summary Tasks it might be possible to create such logic, but not always. It will not require the addition of dummy milestones but usually will hide predecessor logic. This is as worst than the prior option.
Option 5: If you have software capable of modeling Pacing Hammock as defined by Plotnick (http://www.rdcpm.com/07sem-01.pdf), that can display all predecessor logic then you will have no exceptions like those possible under the traditional implementation of Summary Task by MSP, there will be no possibility of circular references, there will be no need for additional milestones that can create confusion. You will need a pacing hammock that will include all predecessors to A, another for all predecessor to B, another for all predecessors to C, and another for unlinked Activities A, B and C, finally just link the Pacing Hammocks.
Unfortunately no single software I know provides for such Pacing Hammock functionality. Not even Primavera Risk that can model Linked Summary Tasks but misses to display all predecessor/successor logic between the member activities of each Summary Task. Ironically Primavera Risk complies with the requisites by Mr. Plotnick with regard to his coding standard but not with regard to the purpose of it, to provide a transparent view on the logic. It provide for "manual (optional) transparency" that must be continuously revised in order to keep it valid. This I like but do not know of a software that implements the Pacing Hammock or its equivalent while at the same time all predecessor/successor logic is visible on a tabular report.
All five options will do the job, it is up to you and your software to decide, I bet Option 3 will be your best available option.