Adaptive predictive analytics for design modification requests

One or more non-transitory computer-readable storage media coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for automating change requests relating to a product manufactured through a supply chain network, the operations comprising:

• receiving training data that includes private information for a node in a supply chain network that is private from other nodes in the supply chain network and information regarding previous decisions related to product change requests for a product manufactured through the supply chain network,
• where each node in the supply chain network represents a respective supplier that supplies a respective part used in producing a product and locations of the nodes in the supply chain network reflect which parts supplied by the suppliers are used by the suppliers to supply other parts;
• training, using the training data, a predictive model for the node that is configured to render decisions that approve or deny requests to change a part used in manufacturing the product;
• after a respective predictive model for each other node in the supply chain network is trained, with respective private data that is private from the other nodes in the supply chain network and the information regarding the previous decisions related to the product change requests for the product manufactured through the supply chain network, to render decisions that approve or deny requests to change a respective part used in manufacturing the product, receiving, from a requesting node in the supply chain network and by the node, a request to change a given part that is supplied by a requesting supplier represented by the requesting node and used in a part supplied by a given supplier represented by the node;
• applying the predictive model for the node to the request to change the given part;
• based on applying the predictive model for the node to the request to change the given part, identifying a plurality of other nodes in the supply chain network to forward the request;
• transmitting the request to change the given part to the other nodes that were identified;
• receiving, from each of the other nodes that were determined, a response to the request to change the given part, wherein each response from the other nodes was determined based on the respective predictive model of the other node and includes an approval or denial of the request to change the given part by the other node;
• applying the predictive model for the node to the responses received from the other nodes;
• based on applying the predictive model for the node to the received responses from the other nodes, determining a decision approving or denying the request to change the given part;
• and
• transmitting the decision to the requesting node.

A system, comprising:

• one or more processors;
• and
• computer-readable storage device coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
• receiving training data that includes private information for a node in a supply chain network that is private from other nodes in the supply chain network and information regarding previous decisions related to product change requests for a product manufactured through the supply chain network,
• where each node in the supply chain network represents a respective supplier that supplies a respective part used in producing a product and locations of the nodes in the supply chain network reflect which parts supplied by the suppliers are used by the suppliers to supply other parts;
• training, using the training data, a predictive model for the node that is configured to render decisions that approve or deny requests to change a part used in manufacturing the product;
• after a respective predictive model for each other node in the supply chain network is trained, with respective private data that is private from the other nodes in the supply chain network and the information regarding the previous decisions related to the product change requests for the product manufactured through the supply chain network, to render decisions that approve or deny requests to change a respective part used in manufacturing the product, receiving, from a requesting node in the supply chain network and by the node, a request to change a given part that is supplied by a requesting supplier represented by the requesting node and used in a part supplied by a given supplier represented by the node;
• applying the predictive model for the node to the request to change the given part;
• based on applying the predictive model for the node to the request to change the given part, identifying a plurality of other nodes in the supply chain network to forward the request;
• transmitting the request to change the given part to the other nodes that were identified;
• receiving, from each of the other nodes that were determined, a response to the request to change the given part, wherein each response from the other nodes was determined based on the respective predictive model of the other node and includes an approval or denial of the request to change the given part by the other node;
• applying the predictive model for the node to the responses received from the other nodes;
• based on applying the predictive model for the node to the received responses from the other nodes, determining a decision approving or denying the request to change the given part;
• and
• transmitting the decision to the requesting node.