Building a algoritm for lighting calibration (from science papers) 10+ Hard

We have a lot of science Papers all with valuable information and formulas, this needs to be analyzed, taken out the needed information and conducted into an algorithm for color calibration. It should do: Take LED binning data. Optical stats Voltage stats Driver efficienty Options for chromatical adaption models Then controlable lumen output and color in k, with extremely high accuracy. So all the variables: Temperature Depreciation Led binning data Fixture data Driver data With ai learning It has to work extremely accurate, with 15 scientific papers that need to be analyzed and taken codes and formulas from. So we need: A algorithm that can calibrate led fixtures based on binning data, with a tolerance of 0.05%. We have all the needed info across 15+ papers with all it’s own formulas written in symbols and physics data, this needs to be translated into algorithms and should end into a contribution value for a driver. For example i want 500 lumen (so optical loss, driver loss, depreciation, fixture loss, current etc) deducted. And 3000k it should calculate the color point for that specific binning. With all the factors