I am planning to estimate timelines by using the bottom up technique of each activity for scientific R&D project.
In your opinion, what is the suitable approach for monitoring it?
Monitoring SPI, or using buffer consumption fever chart by adding 50% buffer?
Why not use a combination of EVM for overall performance tracking (assuming your reporting systems are set up to support proper EVM) and buffer management for monitoring the health of critical and feeder network paths?
I'd also avoid use a flat percentage for buffers - those should be based on the quantified schedule risks associated with each network path.
R&D projects are notoriously difficult to apply EVM. Outside R&D, I often hear the expression, "We don't want to turn this into a big science project." while in R&D that is exactly what you have.
The key in my experience is being able to break the early work down into small enough meaningful pieces. Early in R&D, there are often extended periods of study and experimentation. Tracking tasks that are multiple months long is prone to people just calculating where their SPI = 1.0 and entering that % complete. You don't know you're behind schedule until too late.
To have a meaningful estimate of actual % complete, you need to subdivide those long periods of exploring the unknown into more discrete tasks. This could be things like how many tests you need to run over a time period, development of discrete functions, the gap between required properties and currently achievable properties, etc. It really depends on the field of research.
Without planning the R&D as smaller discrete activities that have clearly defined start and end points, statusing how close you are to discovering the next big breakthrough tends to turn into a game of how to avoid negative attention for falling behind. EVM works best when problems are identified early, so the real focus needs to be on planning the early work, if you are to be successful tracking it. Saving Changes...
There are various forms of estimating Activity Duration: Analogous,
Parametric, Three-Point, Bottom-Up, and also expert judgment.
For cascade projects, you can develop a schedule using MS Project, and monitor using EVM principles and S-Curves implementation, establishing also % completed and physical % completed.
For agile projects, it's adequate to use Jira or Trello software. Also, there is a useful tool that connects Trello with Power BI, allowing to have a management monitoring dashboard. Check it at https://www.facebook.com/Teleworking-Monitoring-107369664431280 Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
I am leading in R&D program/projects. I used EVM in the past. But my recommendation is: going for buffer consumption just in case you are talking about something closer to constraints theory or burn-down which is an implementation. What I try to say is: mainly in R&D projects if you create an schedule things will not happened as the schedule shows except your schedule is at summary tasks level and you manage it by making a relation between total duration/total work/remained duration/remainded work. What I try to say is forget about monitoring the project at detailled activity level. If your R&D project is following the detailled activities as planned then you are in trouble with your R&D project. And by the way, with no aim to include a buzzword here, R&D projects are the best candidates to use agile based methods. We are using Scrum and DSDM. Saving Changes...
Vladimir LiberzonR&D Director| Spider Project TeamMoscow, Russian Federation
R&D projects have a lot of risks and uncertainties. It is necessary to include in the model probabilistic and conditional branches, enter uncertainties of activity parameters, simulate risks and create contingency reserves (project buffers) that are defined not as 50% but as sufficient for reasonable probability of meeting planned targets. Managing projects look what happens with success probability. It is the best measure of buffer consumption. If success probability trend is negative then project buffer is consumed faster than expected and corrective actions may be required.
This is Success Driven Project Management methodology that is proved method of managing projects with high uncertainty. Saving Changes...
Community Champion
