US 12,579,500 B2 · Supply chain good inspection utilizing machine learned robotic process automation

Abstract

A value chain system that provides recommendations for designing a logistics system generally includes a machine learning system that trains machine-learned models that output logistics design recommendations based on training data sets that each respectively defines one or more features of a respective logistic system and an outcome relating to the respective logistics system; an artificial intelligence system that receives a request for a logistics system design recommendation and determines the logistics system design recommendation based on one or more of the machine-learned models and the request; and a digital twin system that generates an environment digital twin of a logistics environment that incorporates the logistics system design recommendation, and one or more physical asset digital twins of physical assets. The digital twin system executes a simulation based on the logistics environment digital twin, the one or more physical asset digital twins.

An information technology system, comprising:

a set of communication systems to enable systems and applications associated with a plurality of value chain network entities to communicate amongst themselves, wherein the plurality of value chain network entities includes at least a smart manufacturing device and a product inspection facility;
a set of monitoring systems that monitor and collect data related to at least one of:
components, devices, environments, or activities associated with a product, wherein each of the set of monitoring systems are communicatively coupled, via at least one of the set of communication systems, with at least a corresponding IoT monitoring system or a corresponding sensor system deployed in the at least the smart manufacturing device and the product inspection facility;
a set of data storage systems to store current data collected from the set of monitoring systems and at least one of:
historical data, historical outcomes, or past behavior data, and information derived therefrom;
a digital twin system to model environments, equipment, processes, and products associated with a subset of the plurality of value chain network entities, wherein the digital twin system enables creation of digital twins including a product digital twin, a smart manufacturing device digital twin, and a product inspection process digital twin;
a set of adaptive intelligence systems to create and train, using a modular neural network including a set of independent neural networks moderated by an intermediary, a set of machine learned models including:
a manufacturing model to relate manufacturing parameters and product characteristics,
wherein at least a first subset of the set of independent neural networks is trained on a training data set including the past behavior data and the historical data of a respective value chain network entity for pattern recognition,
wherein the set of machine learned models uses at least one of:
a specification of the smart manufacturing device, data from maintenance records of the smart manufacturing device, operating data collected from sensors connected or in proximity to the smart manufacturing device, or data from similar smart manufacturing devices, and
wherein the pattern recognition of the manufacturing model includes recognizing one or more relationships between a manufacturing process and the product characteristics;
and
a product inspection model to generate a product quality determination when provided with product data and images,
wherein at least a second subset of the set of independent neural networks is trained on historical information and data stored on the set of data storage systems, including a determined product quality parameter and at least one corresponding set of:
inspector interactions from a video or a still image of an inspection of the product by a human inspector or a set of product data characterizing the product being inspected;
and
a unified set of robotic process automation systems to coordinate automation with respect to the plurality of value chain network entities by:
sharing inputs, data structures, data sources, events, states, outputs, or outcomes;
performing, by one of the unified set of robotic process automation systems, an automatic product inspection process on an in-process product using the product inspection process digital twin, the product inspection model, an image of the product being inspected, and corresponding product data;
determining, by one of the unified set of robotic process automation systems, a product quality flag based on an output of the automatic product inspection process;
upon determination of a product quality flag indicating a defect in the product, determining, by one of the unified set of robotic process automation systems, a correlation between the manufacturing process and the defect in the product, using a simulation of the smart manufacturing device digital twin, the manufacturing model, and corresponding product and process data;
upon determination of a positive correlation between the manufacturing process and the defect, generating an automated control signal based on the output of the smart manufacturing device digital twin;
and
providing the automated control signal to the smart manufacturing device using at least one of the set of communication systems;
and
altering, by the smart manufacturing device, a production rate in response to the automated control signal.

24 more elements

Supply chain good inspection utilizing machine learned robotic process automation

Life of this patent

Supply chain good inspection utilizing machine learned robotic process automation

Life of this patent

Abstract

Claims21