Intelligent dishwashing systems and methods

Abstract

Example intelligent dishwashing systems and methods are described. In one implementation, a system includes an imaging system configured to capture at least one image of an article of dishware. A processing system analyzes the at least one image to determine a presence and a location of a stain on the article of dishware. Responsive to determining the presence of a stain, a cleaning system is configured to clean the location of the stain on the article of dishware.

1 - 27 . (canceled) 
     
     
         28 . A method for adjusting a machine parameter of a dishwashing system, the method comprising:
 obtaining an image of an article of dishware;   classifying food soils on the article of dishware using the image of the article of dishware and a neural network that is trained to classify food soils on articles of dishware based on images of articles of dishware; and   adjusting a machine parameter of a dishwashing system in response to the classification of the food soils on the article of dishware.   
     
     
         29 . The method of  claim 28 , wherein the food soils are classified in terms of freshness. 
     
     
         30 . The method of  claim 28 , wherein the food soils are classified in terms of an amount of food soils. 
     
     
         31 . The method of  claim 28 , wherein the food soils are classified in terms of a type of food soils. 
     
     
         32 . The method of  claim 28 , wherein the machine parameter of a dishwashing system includes at least one of water, power, temperature, chemical dosage, and time. 
     
     
         33 . The method of  claim 28 , further comprising training the neural network by:
 labeling images of articles of dishware with labels that corresponds to a time that was taken to completely clean the respective articles of dishware; and   training the neural network, using the labeled images, to classify images of articles of dishware with a food soil classification.   
     
     
         34 . The method of  claim 33 , further comprising tracking the time that was taken for an article of dishware to be cleaned and using the time as a label for an image of the corresponding article of dishware, wherein the image was obtained before the article of dishware was cleaned. 
     
     
         35 . The method of  claim 33 , further comprising labeling the images of articles of dishware with labels that correspond to knowledge of food items that may be placed in or on the article of dishware and training the neural network, using the labeled images, to classify images of articles of dishware with a food soil classification. 
     
     
         36 . The method of  claim 35 , wherein the food soils on the article of dishware are classified using the image of the article of dishware and the knowledge of food items that may be placed in or on the article of dishware. 
     
     
         37 . The method of  claim 28 , further comprising training the neural network by:
 labeling images of articles of dishware with labels that corresponds to a time that was taken to completely clean the respective articles of dishware; and   training the neural network, using the labeled images, to classify images of articles of dishware with a degree of freshness.   
     
     
         38 . The method of  claim 37 , further comprising tracking the time that was taken for an article of dishware to be cleaned and using the time as a label for an image of the corresponding article of dishware, wherein the image was obtained before the article of dishware was cleaned. 
     
     
         39 . The method of  claim 28 , further comprising labeling the images of articles of dishware with labels that correspond to knowledge of food items that may be placed in or on the article of dishware. 
     
     
         40 . The method of  claim 28 , wherein the neural network is a convolutional neural network (CNN). 
     
     
         41 . An automated dishwashing system comprising:
 a processing system configured to:   obtain an image of an article of dishware;   classify food soils on the article of dishware using the image of the article of dishware and a neural network that is trained to classify food soils on articles of dishware based on images of articles of dishware; and   adjust a machine parameter of a dishwashing system in response to the classification of the food soils on the article of dishware.   
     
     
         42 . A method for operating a dishwashing system, the method comprising:
 obtaining an image of an article of dishware;   estimating how long it would take to clean an article of dishware using the image of the article of dishware and a neural network that is trained to estimate cleaning time based on images of articles of dishware; and   cleaning the article of dishware according to the estimation.   
     
     
         43 . A method for training a neural network, the method comprising:
 labeling images of articles of dishware with labels that correspond to a time that was taken to clean the respective articles of dishware; and   training a neural network, using the labeled images, to classify images of articles of dishware with a food soil classification.   
     
     
         44 . The method of  claim 43 , further comprising tracking the time that was taken for an article of dishware to be cleaned and using the time as a label for an image of the corresponding article of dishware, where the image was obtained before the article of dishware was cleaned. 
     
     
         45 . The method of  claim 43 , wherein the food soils are classified in terms of freshness. 
     
     
         46 . The method of  claim 43 , wherein the food soils are classified in terms of an amount of food soils. 
     
     
         47 . The method of  claim 43 , wherein the food soils are classified in terms of a type of food soils. 
     
     
         48 . The method of  claim 43 , further comprising labeling the images of articles of dishware with labels that correspond to knowledge of food items that may be placed in or on the article of dishware. 
     
     
         49 . The method of  claim 48 , wherein the knowledge includes knowledge of a menu. 
     
     
         50 . The method of  claim 48 , wherein the knowledge includes knowledge of a business. 
     
     
         51 . A method for training a neural network, the method comprising:
 labeling images of articles of dishware with labels that correspond to a time that was taken to clean the respective articles of dishware; and   training a neural network, using the labeled images, to estimate a degree of freshness of articles of dishware.   
     
     
         52 . The method of  claim 51 , further comprising tracking the time that was taken for an article of dishware to be cleaned and using the time as a label for an image of the corresponding article of dishware, where the image was obtained before the article of dishware was cleaned. 
     
     
         53 . The method of  claim 51 , further comprising labeling the images of articles of dishware with labels that correspond to a time that the corresponding article of dishware was dropped off by a user of the article of dishware.