Granted Patent B2

US 12,298,739 · App. 17/661,921 · Granted May 13, 2025

Inventors: Prachin Lalit Jain (Thane, IN); Sanket Junagade (Thane, IN); Sanat Sarangi (Thane, IN); Srinivasu Pappula (Hyderabad, IN)

Assignee: TATA CONSULTANCY SERVICES LIMITED

G05B19/4155G05B2219/37375

Abstract

Claims (45)

1. A processor implemented method of micro-climate management, comprising:

collecting sensor data with respect to ambient conditions and field conditions in a plurality of plots in a land area being monitored, using a plurality of sensors, via one or more hardware sensors, wherein the plurality of sensors comprises a plurality of mobile sensors, and at least one fixed sensor node, wherein the at least one fixed sensor node is associated with a plurality of ambient sensors, wherein the at least one fixed sensor node is placed in a plot P, and wherein the plot P is at a location at an approximate centre of a farm;

grouping all sensors from among the plurality of sensors, that are identified as following same trend in a trend analysis of the plurality of sensors, to generate one or more sensor groups, via the one or more hardware sensors, wherein the trend analysis comprises:

generating at least one dataset, wherein the generated at least one dataset comprising a plurality of data points extracted from the sensor data;

plotting a linear regression curve for the generated at least one dataset;

determining the trend as one of increasing or decreasing, for the ambient conditions in the collected sensor data; and

generating the one or more sensor groups, based on the determined trend,

wherein the one or more sensor groups includes group G1 and/or group G2 generated by performing the trend analysis, plots in the group G1 are similar to the plot P with the same trend as that of the plot P and plots in the group G2 are different from that of the plot P with different trend from the trend of the plot P;

performing regrouping of the plurality of sensors forming the one or more sensor groups to form one or more regrouped sensor groups, by iteratively performing the regrouping based on a determined homogeneity of the plurality of sensors forming the one or more sensor groups, via the one or more hardware sensors, wherein the one or more regrouped sensor groups are formed by performing a homogeneity check of the plurality of sensors forming the one or more sensor groups using a combination of an Analysis of Variance (ANOVA) test and a Spearman Rank Correlation Test techniques, wherein the ANOVA test on the one or more sensor groups to check if the one or more sensor groups are homogeneous within themselves or not and after the ANOVA test is done, resulting p-value is checked to determine if the p-value is exceeding a pre-defined value and if the p-value is identified as below the pre-defined value, then an entire group of plots is not homogeneous and rejects a null hypothesis, otherwise the entire group of plots is homogeneous, wherein when the one or more sensor groups fails the ANOVA test, then the Spearman Rank Correlation Test is performed on the one or more sensor groups to check for homogeneity of each plot individually with all other plots in the one or more sensor groups, wherein a Spearman Rank Correlation is used to perform the Spearman Rank Correlation Test and declare the plots homogeneous;

generating a micro-climate view of the land area, wherein generating the micro-climate view comprises grouping plots from among the plurality of plots, based on the regrouped sensor groups, via the one or more hardware sensors, wherein the plots forming the micro-climate view are homogenous plots, and wherein the micro-climate view indicates different groups of homogeneous plots in a way that the plots in different homogeneous groups have unique characteristics in terms of the ambient conditions and field conditions; and

performing a micro-climate forecast of the ambient conditions in the land area being monitored, for at least one future instance of time, based on the micro-climate view, via the one or more hardware sensors, wherein performing the micro-climate forecast comprises:

determining, using a time series forecasting process, a weekly forecast data using a historical daily plot data for at least one homogenous plot from among a plurality of homogenous plots forming the micro-climate view;

filling missing data for one or more plots that are homogenous with the at least one homogenous plot for which the weekly forecast data has been determined, as forecast data; and

generating one or more predictions based on the determined forecast data.

2. A system for micro-climate management, comprising:

one or more hardware processors;

an I/O interface; and

a memory storing a plurality of instructions, wherein the plurality of instructions when executed, cause the one or more hardware processors to:

collect sensor data with respect to ambient conditions and field conditions in a plurality of plots in a land area being monitored, using a plurality of sensors, wherein the plurality of sensors comprises a plurality of mobile sensors, and at least one fixed sensor node, wherein the at least one fixed sensor node is associated with a plurality of ambient sensors, wherein the at least one fixed sensor node is placed in a plot P, and wherein the plot P is at a location at an approximate centre of a farm;

group all sensors from among the plurality of sensors, that are identified as following same trend in a trend analysis of the plurality of sensors, to generate one or more sensor groups, wherein the trend analysis comprises:

generating at least one dataset, wherein the generated at least one dataset comprising a plurality of data points extracted from the sensor data;

plotting a linear regression curve for the generated at least one dataset;

determining the trend as one of increasing or decreasing, for the ambient conditions in the collected sensor data; and

generating the one or more sensor groups, based on the determined trend,

wherein the one or more sensor groups includes group G1 and/or group G2 generated by performing the trend analysis, plots in the group G1 are similar to the plot P with the same trend as that of the plot P and plots in the group G2 are different from that of the plot P with different trend from the trend of the plot P;

perform regrouping of the plurality of sensors forming the one or more sensor groups to form one or more regrouped sensor groups, by iteratively performing the regrouping based on a determined homogeneity of the plurality of sensors forming the one or more sensor groups, wherein the one or more regrouped sensor groups are formed by performing a homogeneity check of the plurality of sensors forming the one or more sensor groups using a combination of an Analysis of Variance (ANOVA) test and a Spearman Rank Correlation Test techniques, wherein the ANOVA test on the one or more sensor groups to check if the one or more sensor groups are homogeneous within themselves or not and after the ANOVA test is done, resulting p-value is checked to determine if the p-value is exceeding a pre-defined value and if the p-value is identified as below the pre-defined value, then an entire group of plots is not homogeneous and rejects a null hypothesis, otherwise the entire group of plots is homogeneous, wherein when the one or more sensor groups fails the ANOVA test, then the Spearman Rank Correlation Test is performed on the one or more sensor groups to check for homogeneity of each plot individually with all other plots in the one or more sensor groups, wherein a Spearman Rank Correlation is used to perform the Spearman Rank Correlation Test and declare the plots homogeneous;

generate a micro-climate view of the land area, wherein generating the micro-climate view comprises grouping plots from among the plurality of plots, based on the regrouped sensor groups, wherein the plots forming the micro-climate view are homogenous plots, and wherein the micro-climate view indicates different groups of homogeneous plots in a way that the plots in different homogeneous groups have unique characteristics in terms of the ambient conditions and field conditions; and

perform a micro-climate forecast of the ambient conditions in the land area being monitored, for at least one future instance of time, based on the micro-climate view, wherein performing the micro-climate forecast comprises:

determining, using a time series forecasting process, a weekly forecast data using a historical daily plot data for at least one homogenous plot from among a plurality of homogenous plots forming the micro-climate view;

filling missing data for one or more plots that are homogenous with the at least one homogenous plot for which the weekly forecast data has been determined, as forecast data; and

generating one or more predictions based on the determined forecast data.

3. One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

collecting sensor data with respect to ambient conditions and field conditions in a plurality of plots in a land area being monitored, using a plurality of sensors, wherein the plurality of sensors comprises a plurality of mobile sensors, and at least one fixed sensor node, wherein the at least one fixed sensor node is associated with a plurality of ambient sensors, wherein the at least one fixed sensor node is placed in a plot P, and wherein the plot P is at a location at an approximate centre of a farm;

grouping all sensors from among the plurality of sensors, that are identified as following same trend in a trend analysis of the plurality of sensors, to generate one or more sensor groups, wherein the trend analysis comprises:

generating at least one dataset, wherein the generated at least one dataset comprising a plurality of data points extracted from the sensor data;

plotting a linear regression curve for the generated at least one dataset;

determining the trend as one of increasing or decreasing, for the ambient conditions in the collected sensor data; and

generating the one or more sensor groups, based on the determined trend,

wherein the one or more sensor groups includes group G1 and/or group G2 generated by performing the trend analysis, plots in the group G1 are similar to the plot P with the same trend as that of the plot P and plots in the group G2 are different from that of the plot P with different trend from the trend of the plot P;

performing regrouping of the plurality of sensors forming the one or more sensor groups to form one or more regrouped sensor groups, by iteratively performing the regrouping based on a determined homogeneity of the plurality of sensors forming the one or more sensor groups, wherein the one or more regrouped sensor groups are formed by performing a homogeneity check of the plurality of sensors forming the one or more sensor groups using a combination of an Analysis of Variance (ANOVA) test and a Spearman Rank Correlation Test techniques, wherein the ANOVA test on the one or more sensor groups to check if the one or more sensor groups are homogeneous within themselves or not and after the ANOVA test is done, resulting p-value is checked to determine if the p-value is exceeding a pre-defined value and if the p-value is identified as below the pre-defined value, then an entire group of plots is not homogeneous and rejects a null hypothesis, otherwise the entire group of plots is homogeneous, wherein when the one or more sensor groups fails the ANOVA test, then the Spearman Rank Correlation Test is performed on the one or more sensor groups to check for homogeneity of each plot individually with all other plots in the one or more sensor groups, wherein a Spearman Rank Correlation is used to perform the Spearman Rank Correlation Test and declare the plots homogeneous;

generating a micro-climate view of the land area, wherein generating the micro-climate view comprises grouping plots from among the plurality of plots, based on the regrouped sensor groups wherein the plots forming the micro-climate view are homogenous plots, and wherein the micro-climate view indicates different groups of homogeneous plots in a way that the plots in different homogeneous groups have unique characteristics in terms of the ambient conditions and field conditions; and

performing a micro-climate forecast of the ambient conditions in the land area being monitored, for at least one future instance of time, based on the micro-climate view, wherein performing the micro-climate forecast comprises:

determining, using a time series forecasting process, a weekly forecast data using a historical daily plot data for at least one homogenous plot from among a plurality of homogenous plots forming the micro-climate view;

filling missing data for one or more plots that are homogenous with the at least one homogenous plot for which the weekly forecast data has been determined, as forecast data; and

generating one or more predictions based on the determined forecast data.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (29)