Customizing agricultural practices to maximize crop yield

Patent metadata

Field	Value
Publication number	US-11645308-B2
Application number	US-202016884776-A
Country	US
Kind code	B2
Filing date	May 27, 2020
Priority date	May 27, 2020
Publication date	May 9, 2023
Grant date	May 9, 2023

Abstract

Methods, systems, and computer program products for customizing agricultural practices to maximize crop yield are provided herein. A computer-implemented method includes obtaining data pertaining to (i) a geographical area comprising a plurality of regions and (ii) one or more agricultural practices applied to the geographical area; assigning each of the plurality of regions to a respective cluster of a set clusters, based at least in part on comparing features identified in the data, wherein similar ones of said regions are assigned to the same cluster; generating instructions that are specific to a given cluster in the set, wherein the instructions relate to agricultural tasks to be performed on the regions assigned to the given cluster; and triggering, based on said instructions, one or more automated farming processing devices, thereby carrying out at least a portion of said agricultural tasks.

First claim

What is claimed is:

1. A computer-implemented method, the method comprising:
obtaining data pertaining to a geographical area comprising a plurality of regions and one or more agricultural practices applied to the geographical area;
providing the obtained data to a machine learning regressor model, wherein the machine learning regressor model is trained to predict a crop yield for each of the plurality of regions based on a plurality of features identified in the data;
applying a machine learning model to determine a relative impact of each of the plurality of features on the crop yield predicted by the machine learning regressor model for each of the plurality of regions, wherein the relative impact of each of the features corresponds to one of: a negative impact value and a positive impact value;
identifying a subset of the plurality of features having the greatest impact on the predicted crop yield;
applying an unsupervised clustering process to assign each of the plurality of regions to a respective cluster of a set clusters, wherein the unsupervised clustering process assigns similar ones of said regions to the same cluster based at least on the subset of the plurality of features;
generating instructions that are specific to a given cluster in the set, wherein the instructions relate to agricultural tasks to be performed on the regions assigned to the given cluster, wherein said generating comprises: constructing a causal graph for the given cluster comprising a plurality of nodes and a plurality of edges, wherein each node represents one of the plurality of regions and its corresponding data, and each edge represents distance between the regions of the nodes corresponding to the edge and one or more transportation conditions between the regions of the nodes corresponding to the edge, wherein the one or more transportation conditions correspond to one or more geographical features between the regions;
transmitting said instructions to one or more automated farming processing devices in the form of computer code, thereby triggering the one or more automated farming devices to perform at least a portion of said agricultural tasks; and
updating over time at least one of: the set of clusters and the causal graph based on performance of said agricultural tasks;
wherein the method is carried out by at least one computing device.

2. The computer-implemented method of claim 1 , wherein said generating comprises:
generating the instructions to increase the crop yield for regions assigned to said given cluster.

3. The computer-implemented method of claim 1 , comprising:
in response to said transmitting, obtaining further data from one or more sensors of said one or more automated farming processing devices; and
adjusting said instructions based at least in part on said further data.

4. The computer-implemented method of claim 1 , comprising:
obtaining information from one or more subject matter experts pertaining to one or more of the plurality of regions, in response to one or more counterfactual queries for the given cluster.

5. The computer-implemented method of claim 4 , comprising:
training the machine learning regressor model based at least on (i) historical crop yields for said geographical area and historical agricultural practices for said geographical area; and
generating said counterfactual queries using the trained machine learning regressor model.

6. The computer-implemented method of claim 1 , comprising:
obtaining one or more constraints for generating said instructions, the one or more constraints comprising one or more of: (i) maximizing an overall crop yield for the given cluster; minimizing changes to existing agricultural practices; and decreasing the number of clusters.

7. The computer-implemented method of claim 1 , wherein said generating said instructions comprises updating said causal graph over time based on performance of said agricultural tasks.

8. The computer-implemented method of claim 1 , wherein the data comprise one or more of:
global positioning data;
weather data;
data indicative of one or more characteristics for one or more crops;
data indicative of soil moisture;
data from one or more remote sensors associated with said geographical area; and
data indicative of soil moisture levels.

9. The computer-implemented method of claim 1 , wherein:
for each respective one of the features, the corresponding impact value is output by the machine learning model to form an impact value space; and
the unsupervised clustering process is applied on at least a portion of the impact value space to assign each of the plurality of regions.

10. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:
obtain data pertaining to (i) a geographical area comprising a plurality of regions and one or more agricultural practices applied to the geographical area;
provide the obtained data to a machine learning regressor model, wherein the machine learning regressor model is trained to predict a crop yield for each of the plurality of regions based on a plurality of features identified in the data;
apply a machine learning model to determine a relative impact of each of the plurality of features on the crop yield predicted by the machine learning regressor model for each of the plurality of regions, wherein the relative impact of each of the features corresponds to one of: a negative impact value and a positive impact value;
identify a subset of the plurality of features having the greatest impact on the predicted crop yield;
apply an unsupervised clustering process to assign each of the plurality of regions to a respective cluster of a set clusters, wherein the unsupervised clustering process assigns similar ones of said regions to the same cluster based at least on the subset of the plurality of features;
generate instructions that are specific to a given cluster in the set, wherein the instructions relate to agricultural tasks to be performed on the regions assigned to the given cluster, wherein said generating comprises: constructing a causal graph for the given cluster comprising a plurality of nodes and a plurality of edges, wherein each node represents one of the plurality of regions and its corresponding data, and each edge represents distance between the regions of the nodes corresponding to the edge and one or more transportation conditions between the regions of the nodes corresponding to the edge, wherein the one or more transportation conditions correspond to one or more geographical features between the regions;
transmit said instructions to one or more automated farming processing devices in the form of computer code, thereby triggering the one or more automated farming devices to perform at least a portion of said agricultural tasks; and
update over time at least one of: the set of clusters and the causal graph based on performance of said agricultural tasks.

11. The computer program product of claim 10 , wherein said generating comprises:
generating the instructions to increase the crop yield for regions assigned to said given cluster.

12. The computer program product of claim 10 , wherein the program instructions executable by a computing device further cause the computing device to:
in response to said transmission, obtaining further data from one or more sensors of said one or more automated farming processing devices; and
adjusting said instructions based at least in part on said further data.

13. The computer program product of cla

Assignees

• IBM

Inventors

• MARVANIYA Smitkumar Narotambhai
• Godbole Shantanu R
• Mukherjee Sumanta
• Raykar Vikas C

Classifications

• A01B79/005 Primary

  Precision agriculture · CPC title

• G06F16/2458

  Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries · CPC title

• G06F16/29

  Geographical information databases · CPC title

• G06F16/285 Primary

  Clustering or classification · CPC title

• G06Q50/02

  Agriculture; Fishing; Forestry; Mining · CPC title