Hallo. Ich hoffe, Ihnen geht es gut. I checked out the attached file and I think I get what you need. It is fairly complicated, but I'm very interested to work on this. I guess that a good way to do it would be start defining what each innovation is, how will it affect the output (put a weight to each innovation), but more importantly, since this is non-linear, is how each innovation would affect the other: for example, what if innovation 1 provides 50% improvement, and innovation 2 provides another 50%, (you would expect the whole improvement to be 75%, since it's x * 0.5 * 0.5) but what if combined they only make 40%-60%? I guess once you define your innovations, any kind of model could be done. But I think the best way to go is by assigning weights to each innovation.