Automatic sorting of waste

Abstract

A waste sorting system uses various sensors to automatically determine whether a piece of waste is recyclable, and to dispense the waste into an appropriate waste bin of a multi-bin device, at least one of which is a recyclable waste bin. The system also uses sensors to determine a type of the waste, and it causes a display device or speaker to output information about the type of waste.

1 . A method of automatically sorting waste, the method comprising, by a processor:
 detecting that a system has received a piece of waste, wherein the system comprises a first waste bin and a second waste bin, wherein the first waste bin is a recyclable waste bin;   receiving, from a plurality of sensors of the system, a plurality of properties associated with the piece of waste, wherein the sensors comprise at least two different sensor types;   using at least one of the properties from at least a first one of the sensors to identify a type of the piece of waste, and to determine whether the piece of waste is formed of a recyclable material;   accessing a data set to retrieve information regarding the type of the piece of waste;   causing a display device or a speaker to output the information; and   in response to determining that the piece of waste is formed of a recyclable material, causing the system to deposit the piece of waste into the recyclable waste bin.   
     
     
         2 . The method of  claim 1  further comprising, by the processor:
 in response to determining that the piece of waste is formed of a recyclable material, determining whether the piece of waste meets a recyclability threshold; and 
 wherein causing the system to deposit the piece of waste into the recyclable waste bin is also performed in response to the piece of waste meeting the recyclability threshold. 
 
     
     
         3 . The method of  claim 2 , wherein determining whether the piece of waste meets the recyclability threshold comprises:
 receiving data from a load sensor of the system; and   using the data to determine whether the piece of waste meets or exceeds a weight threshold.   
     
     
         4 . The method of  claim 1 , wherein:
 a first one of the sensors comprises a camera; and   one of more additional sensors of the sensors comprises an inductance sensor, a capacitance sensor, a load sensor, a temperature sensor, an infrared light sensor, a near-infrared light sensor, a spectral measurement device, an audio sensor, a capacitive sensor, a fluorescence sensor, or a millimeter-wave radar sensors;   the method further comprises using the camera to capture an image of the piece of waste; and   determining whether the piece of waste is formed of a recyclable material comprises:
 applying a machine learning model to the image to determine a probability that the piece of waste is formed of a recyclable material or a non-recyclable material, and 
 weighting the probability determined from the image more heavily than data received from each of the one of more additional sensors to determine whether the piece of waste is formed of a recyclable material. 
   
     
     
         5 . The method of  claim 4 , wherein applying the machine learning model to the image to determine a probability that the piece of waste is formed of a recyclable material comprises:
 processing the image to determine whether content of the image corresponds to one or more characteristics of a known recyclable object of a data set of known recyclable objects.   
     
     
         6 . The method of  claim 5 , wherein processing the image to determine whether content of the image corresponds to one or more characteristics of a known recyclable object of the data set of known recyclable objects comprises determining whether the content of the image includes one or more of the following:
 a logo of a known recyclable object;   text of a product name of a known recyclable object;   a shape of a known recyclable object; or   a color scheme of a known recyclable object.   
     
     
         7 . The method of  claim 1 , further comprising, based on the type, selecting the recyclable waste bin from a set of candidate bins that include two or more of the following:
 a recyclable plastics bin;   a recyclable paper bin;   a recyclable metal bin;   a recyclable cloth bin;   a recyclable cardboard bin; or   a recyclable glass bin.   
     
     
         8 . The method of  claim 1 , wherein:
 a first one of the sensors comprises an inductance sensor;   a second one of the sensors comprises a load sensor;   determining whether the piece of waste is formed of a recyclable material comprises determining whether the inductance sensor indicates that the piece of waste is metallic;   determining the type of the piece of waste comprises determining that the type is a metallic food or beverage can;   the method also includes determining whether the piece of waste meets or exceeds a recyclability threshold by using data received from the load sensor to determine whether the can meets or exceeds a weight threshold; and   wherein causing the system to deposit the piece of waste into the recyclable waste bin is also performed in response to the piece of waste meeting the recyclability threshold.   
     
     
         9 . The method of  claim 1 , wherein causing the system to deposit the piece of waste in the recyclable waste bin comprises using an actuated sorting department to release the piece of waste to the recyclable waste bin. 
     
     
         10 . A waste container, comprising:
 a waste sorting compartment;   a plurality of waste bins, at least one of which comprises a recyclable waste bin;   a plurality of sensors that comprise at least two different sensor types;   a user interface;   a processor; and   a computer-readable medium containing programming instructions that are configured to instruct the processor to:
 detect a piece of waste in the waste sorting compartment, 
 receive, from the plurality of sensors of the system, a plurality of properties associated with the piece of waste, 
 use at least one of the properties from at least a first one of the sensors to determine a type of the piece of waste, and to determine whether the piece of waste is formed of a recyclable material, 
 access a data set to retrieve information regarding the type of the piece of waste, 
 cause a display device or a speaker to output the information, and 
 in response to determining that the piece of waste is formed of a recyclable material, cause the system to deposit the piece of waste into the recyclable waste bin. 
   
     
     
         11 . The waste container of  claim 10 , wherein the waste container is equipped with the display device or the speaker. 
     
     
         12 . The waste container of  claim 10 , wherein the display device or the speaker is a component of a device that is in communication with the waste container. 
     
     
         13 . The waste container of  claim 10 , further comprising additional programming instructions that are configured to instruct the processor to:
 in response to determining that the piece of waste is formed of a recyclable material, determine whether the piece of waste meets a recyclability threshold; and   wherein the instructions to cause the system to deposit the piece of waste into the recyclable waste bin comprise instructions to do so also in response to the piece of waste meeting the recyclability threshold.   
     
     
         14 . The waste container of  claim 13 , wherein the instructions to determine whether the piece of waste meets the recyclability threshold comprise instructions to:
 use data from a load sensor of the plurality of sensors to determine whether the piece of waste meets or exceeds a weight threshold.   
     
     
         15 . The waste container of  claim 10 , wherein:
 a first one of the sensors comprises a camera; and   one of more additional sensors of the sensors comprises an inductance sensor, a capacitance sensor, a load sensor, a temperature sensor, an infrared light sensor, a near-infrared light sensor, a spectral measurement device, an audio sensor, a capacitive sensor, a fluorescence sensor, or a millimeter-wave radar sensors;   the programming instructions that are configured to cause the processor to determine whether the piece of waste is formed of a recyclable material comprise instructions to:
 apply a machine learning model to an image of the piece of waste captured by the camera to determine a probability that the piece of waste is formed of a recyclable material or a non-recyclable material, and 
 weight the probability determined from the image more heavily than data received from each of the one of more additional sensors to determine whether the piece of waste is formed of a recyclable material. 
   
     
     
         16 . The waste container of  claim 15 , wherein the instructions to apply the machine learning model to the image to determine a probability that the piece of waste is formed of a recyclable material comprise instructions to:
 process the image to determine whether content of the image corresponds to one or more characteristics of a known recyclable object of a data set of known recyclable objects.   
     
     
         17 . The waste container of  claim 16 , wherein the instructions to process the image to determine whether content of the image corresponds to one or more characteristics of a known recyclable object of the data set of known recyclable objects comprise instructions to determine whether the content of the image includes one or more of the following:
 a logo of a known recyclable object;   text of a product name of a known recyclable object;   a shape of a known recyclable object; or   a color scheme of a known recyclable object.   
     
     
         18 . The waste container of  claim 10 , wherein:
 the plurality of waste bins comprise a set of candidate recyclable waste bins that include at least two of the following:
 a recyclable plastics bin, 
 a recyclable paper bin, 
 a recyclable metal bin, 
 a recyclable cloth bin, 
 a recyclable cardboard bin, or 
 a recyclable glass bin; and 
   the programming instructions are further configured to instruct the processor to use the type to select, from the set of candidate bins, the recyclable waste bin into which the piece of waste will be deposited.   
     
     
         19 . The waste container of  claim 10 , wherein:
 a first one of the sensors comprises an inductance sensor;   a second one of the sensors comprises a load sensor;   the instructions to determine whether the piece of waste is formed of a recyclable material comprises determining that the piece of waste is formed of a recyclable material when the inductance sensor indicates that the piece of waste is metallic and type of the piece of waste is a food or beverage can;   programming instructions also include instructions to determine whether the can meets or exceeds a recyclability threshold by using data received from the load sensor to determine whether the can meets or exceeds a weight threshold; and   wherein causing the system to deposit the piece of waste into the recyclable waste bin is also performed in response to the can meeting the recyclability threshold.   
     
     
         20 . The waste container of  claim 10 , further comprising an actuated sorting department that is configured to release the piece of waste to the recyclable waste bin.