Real World Research Progress Curves over Time
I wanted to just throw this out there for technical researchers to be aware of in their own projects based on my experience over several "New to the world" research programs I was a part of during my technical career. Perhaps someone will find it helpful since I have not seen anything really about this in any articles/books on technical research/project management.
Actual Research Progress curves for various target performance characteristics over time for a "New to the world" product program (Graph A):
Three types of progress curves were typically seen in 'New to the world' product development (in my experience).
Typical Research Progress curves over time (Graph B):
These were:
1.Linear progress over time of various slopes. (critical performance characteristics #4,6 in graph A above)
2.Log or Power curve progress over time. (critical performance characteristic #1 in graph A above)
3.Exponential curve progress over time. (critical performance characteristics #2,3,7 in graph A above)
I found knowing which type of progress curve a particular target performance goal seemed to be on (by plotting % attainment of the key performance characteristics over time) was very useful for staffing purposes and predicting completion dates.
For instance, to complete the work necessary to hit 100% of critical performance characteristic #1 in the graph above, additional resources would be needed to help bring that attainment curve back to a linear improvement over time and therefore shorten the total time to reach 100% of the target.
For critical performance characteristics #4,6 - having additional resources added should increase the slope of essentially linear progress toward attainment to get to 100% of the target quicker.
And for critical performance characteristics like #2,3,7 - it isn't obvious how additional resources would necessarily help get to a 100% attainment in a quicker timeframe (since very little or no improvement is initially, or even for quite some time, seen) and in these cases usually, an invention was needed (something not known or predictable beforehand) before any significant improvement was seen (in the project shown above, actually 4 inventions were needed and perhaps even 2 more were going to be needed to make this project successful). However, my approach to critical performance characteristics like #2,3,7 were to at least have enough resources to try out a multitude of different ideas in the hope a positive signal could be seen on a promising pathway - where then additional focus and/or resources could be brought to bear.
Finally, closer inspection of several of the performance characteristics show that sometimes an increase in attainment toward the target in one area could mean a decrease in performance attainment in a different area - even a huge decrease in performance in some cases - BUT several times, the loss in the one area (at the expense of another attribute getting higher performance) could be rather quickly be made up at a rate faster than the initial performance attainment rate and without giving up the initial improvement in the other area.
note - I have always looked at this effect of gaining in one performance area while losing (at least initially) in another performance area as a "how high are various balls in the air being juggled" or "how fast are spinning plates balancing on poles going so they don't fall off" trade off(s) every researcher/product developer has experienced in their own research efforts.
By showing that in some cases the rate of attainment of various performance characteristics can be simulated with known curve shapes, I hoped to give the product or project leader or manager a new tool for better staffing decisions and timeline prediction capabilities
I would appreciate any comments, additions or questions to this concept I've tried to communicate above. Thanks.