End to end platform to manage circular economy of waste materials

1 . A method for dynamically configuring a chemical recycling process, the method comprising:

determining characterization data from a plurality of feedstocks received at a materials recycling distribution facility by performing one or more analytical methods on the plurality of feedstocks, wherein individual feedstocks of the plurality of feedstocks comprise a plurality of waste materials;

accessing the characterization data of a feedstock of the plurality of feedstocks, the characterization data comprising one or more spectral data sets of the feedstock from a plurality of spectral data sets stored within a spectrum database in a computer network, wherein the one or more spectral data sets are collected according to one or more spectroscopic methods;

predicting, using the characterization data, a material composition of the feedstock, wherein the material composition identifies a predicted set of constituent materials;

identifying, at least in part using the predicted material composition of the feedstock, one or more target products based on a set of molecular structural data for the predicted set of constituent materials contained within a set of chemical fingerprint data, wherein the set of molecular structural data is identified using a machine learning model, wherein the machine learning model is trained by using a supervised learning regime or by using sub-band data derived from the one or more spectrum data;

generating a set of candidate chemical reaction schemas enabling a conversion of at least part of the feedstock into the one or more target products, wherein the set of candidate chemical reaction schemas are generated by the machine learning model;

simulating a set of candidate unit operations represented as a set of coupled reaction models, wherein the set of candidate unit operations implements each candidate chemical reaction schema of the set of candidate chemical reaction schemas to convert the predicted material composition of the feedstock into the one or more target products;

directing the feedstock to one or more materials recycling facility destinations;

decomposing the feedstock to the constituent materials at the one or more materials recycling facility destinations operational to perform at least one unit operation from the set of candidate unit operations;

converting the constituent materials to the one or more target products at the one or more materials recycling facility destinations operational to perform at least one unit operation from the set of candidate unit operations,

wherein the one or more materials recycling facility destinations is selected by performing an optimization of the set of candidate chemical reaction schemas to produce a set of optimized chemical reaction schemas, wherein the optimization is based in part on a determination that an output of a reward function exceeds a pre-defined threshold value, and wherein the optimization is also based in part on one or more economic factors, wherein data representing the one or more economic factors comprise a set of utilization data comprising:

a set of market demand data for the one or more target products;

a set of inventory data describing an availability of one or more inputs to the set of candidate chemical reaction schemas;

a set of logistical data comprising a set of recycling feedstock availability data and warehouse storage availability data; and

a set of real time data comprising a set of status data for one or more available recycling facilities to produce the one or more target products.

2 . The method of claim 1 , further comprising:

identifying one or more inputs to a fitness function, the one or more inputs describing the optimized chemical reaction schema of the set of candidate chemical reaction schemas;

generating an output of the fitness function using the one or more inputs; and

selecting an implementation scheme from the set of candidate chemical reaction schemas according to the output of the fitness function, the one or more inputs, and the one or more target products.

3 . The method of claim 2 , wherein the set of candidate chemical reaction schemas are optimized in real time by a computer system to assess the one or more materials recycling facility destinations for recycling the feedstock, wherein the one or more materials recycling facility destinations are determined by one or more fitness values generated by the fitness function implemented by an optimization engine, wherein the one or more fitness values are included as an output by the optimization engine.

4 . The method of claim 1 , wherein identifying the one or more target products comprises:

accessing a set of inventory information data describing a set of products; and

identifying a subset of the set of products as the one or more target products using the set of inventory information data.

5 . The method of claim 1 , wherein predicting the set of constituent materials included in the feedstock comprises:

accessing a library of the spectrum data and associated metadata corresponding to the one or more spectroscopic methods, wherein the library of the spectrum data is stored in a spectrum database that is stored within the computer network;

identifying a band in one or more spectra of the spectrum data using the first machine learning model; and

matching the band to a spectrum of the library of the spectrum data to predict a constituent material of the set of constituent materials using the first machine learning model.

6 . The method of claim 1 , wherein generating the set of candidate chemical reaction schemas comprises:

accessing a chemical reaction inventory comprising representations of chemical reactions describing the conversion of the feedstock to a target product of the one or more target products, wherein the chemical reaction inventory is stored in a chemical inventory database that is stored within the computer network; and

populating the set of candidate of chemical reaction schemas from the chemical reaction inventory, wherein the set of candidate chemical reaction schemas is stored in a chemical reaction schema database that is stored with the computer network.

7 . The method of claim 1 , wherein the reward function is provided with one or more inputs and outputs of the simulation of each chemical reaction schema of the set of candidate chemical reaction schemas, wherein the optimization is based at least in part on iteratively refining the one or more inputs of the simulation to obtain an optimized output of the simulation, wherein the one or more inputs to the simulation is iteratively incremented such that the one or more outputs of the simulation is iteratively changed until the output of the reward function exceeds the pre-defined threshold value, and wherein the one or more inputs to the simulation comprise one or more reaction conditions, one or more catalysts, or one or more reaction models.

8 . The method of claim 1 , wherein the machine learning model includes a convolutional neural network trained on the dataset of molecular structural patterns and spectroscopic data.

9 . The method of claim 1 , wherein the machine learning model includes a reinforcement learning model configured to simulate chemical reactions, optimize target product yield or reduce byproduct generation.

10 . The method of claim 1 , wherein the reward function receives any of yield data, selectivity data, pressure data, temperature data and catalyst level data from the candidate unit operations simulations.

11 . The method of claim 1 , wherein decomposing the feedstock to the constituent materials comprises depolymerizing the feedstock to the constituent materials.

12 . The method of claim 11 , wherein depolymerizing the feedstock comprises pyrolyzing the feedstock to the constituent materials.

13 . The method of claim 1 , wherein performing the optimization of the set of candidate chemical reaction schemas comprises performing a first optimization of the set of candidate chemical reaction schemas based on the determination that the output of the reward function exceeds a pre-defined threshold value, and performing a second optimization of the set of candidate chemical reaction schemas based on the one or more economic factors.

14 . The method of claim 1 , wherein the set of status data is a first set of status data, and a second set of status data for one or more recycling facilities of the one or more target products.

15 . The method of claim 1 , further comprising triggering a redirection of the feedstock from a recipient destination to another recipient destination based on the optimized set of candidate chemical reaction schemas.

16 . The method of claim 1 , wherein the set of candidate reaction schemas combine the machine learning model with first-principles models based on chemical rate equations.