Sorting of materials
AssigneeSortera Technologies, Inc.
InventorsNalin Kumar, Manuel Gerardo Garcia, JR.
Life of this patent
Estimating the term…
Abstract
A system classifies and sorts mixed materials utilizing a combination of sensor systems and/or a vision system that can implement an artificial intelligence system in order to identify or classify each of the materials, which are then sorted into separate groups based on such an identification or classification. The system is capable of classifying and sorting between materials typically found within Zorba, Zebra, and Twitch.
Claims38
A system for sorting material pieces from a stream of materials, comprising:
- a vision system configured to perform one or more vision checks on each material piece within the stream of materials, wherein each of the one or more vision checks comprises classifying each material piece as a function of processing visual images captured from each material piece through an artificial intelligence (“AI”) system;
- a sensor system configured to perform one or more sensor system classifications on each material piece within the stream of materials;
- and
- a sorting apparatus configured to sort material pieces from the stream of materials into one or more classification groups in response to a combination of the one or more vision checks and the one or more sensor system classifications.
A system for sorting of material pieces from a stream of materials, comprising:
- a vision system configured to perform vision checks on each material piece within the stream of materials, wherein each of the vision checks comprises classifying each material piece as a function of processing visual images captured from each material piece through an artificial intelligence (“AI”) system, wherein the vision checks are performed by the vision system implementing a different AI model within the AI system for each of the vision checks;
- an x-ray fluorescence (“XRF”) sensor system configured to perform XRF sensor system classifications on each material piece within the stream of materials, wherein each of the XRF sensor system classifications is performed by a different algorithm analyzing XRF spectral data collected from each material piece by the XRF sensor system;
- and
- a sorting apparatus configured to sort material pieces from the stream of materials as follows:
- a) a plurality of different non-ferrous metal material pieces into separately sorted classification groups based on a first combination of one or more vision checks and one or more sensor system classifications;
- b) a plurality of different cast aluminum alloy material pieces into separately sorted classification groups based on a second combination of one or more vision checks and one or more sensor system classifications;
- and
- c) a plurality of different wrought aluminum alloy pieces material pieces into separately sorted classification groups based on a third combination of one or more vision checks and one or more sensor system classifications.
A computer program product stored on a computer readable storage medium, which when executed performs a method, comprising:
- receiving image data captured by a vision system from each material piece of a mixture of materials moving in a stream past the vision system;
- performing one or more vision checks on each material piece within the stream of materials, wherein each of the one or more vision checks comprises classifications of each material piece as a function of processing the image data captured from each material piece through an artificial intelligence (“AI”) system;
- receiving spectral data captured by a sensor system from each material piece of the mixture of materials moving in the stream past the sensor system;
- performing one or more sensor system classifications on each material piece within the stream of materials as a function of the spectral data;
- and
- sending to a sorting apparatus information regarding the classifications so that the sorting apparatus can sort the material pieces from the stream of materials into one or more classification groups as a function of a combination of the one or more vision checks and the one or more sensor system classifications.
Dependencies are inferred from the claim text.