Abstract: A method includes obtaining plant-related information associated with a growing area, where the plant-related information includes information identifying multiple weeds detected within the growing area. The method also includes processing at least some of the plant-related information to estimate at least one location at risk of weed germination in the growing area. Processing at least some of the plant-related information includes estimating the at least one location at risk of weed germination based at least partially on locations where the identified weeds were detected within the growing area. Estimating the at least one location at risk of weed germination may include performing clustering based on the locations where the identified weeds were detected within the growing area, such as by performing the clustering using a machine learning clustering algorithm.

Abstract: A method includes obtaining, using at least one processing device (202, 314) of an electronic device (102a-102d, 106, 200, 300), spatial information (500, 600, 800, 900, 1100, 1200) associated with weeds in a growing area. The spatial information includes spatial information (500, 800, 1100) associated with a primary weed in the growing area over time. The method also includes identifying, using the at least one processing device, one or more portions of the growing area in which the primary weed has developed or may be developing herbicide resistance using the spatial information. In addition, the method includes outputting, using the at least one processing device, information (700, 1000, 1300) associated with the one or more identified portions of the growing area.

Abstract: A method includes obtaining data measurements associated with plants in at least one growing area. The data measurements are associated with one or more characteristics of the plants and one or more characteristics of the at least one growing area. The method also includes processing at least some of the data measurements to identify one or more anomalous plant production issues associated with the plants. At least one anomalous plant production issue is associated with uneven production by the plants in at least part of the at least one growing area. The method further includes generating at least one visualization presenting at least some of the data measurements or processed versions of at least some of the data measurements in a spatial manner that corresponds to the at least one growing area. The at least one visualization identifies information associated with the one or more anomalous plant production issues.

Abstract: An apparatus includes at least one processor configured to obtain stereo-spatio-temporal data measurements of plants in a growing area. The stereo-spatio-temporal data measurements include (i) first spatio-temporal data measurements of the plants in the growing area and (ii) second spatio-temporal data measurements of the plants in the growing area. The at least one processor is also configured to analyze the stereo-spatio-temporal data measurements to identify one or more actual or potential problems associated with one or more of the plants. The at least one processor is further configured to generate a graphical user interface identifying at least one of the one or more actual or potential problems with the one or more plants. The first and second spatio-temporal data measurements of each stereo-spatio-temporal data measurement are associated with at least one common plant characteristic and different three-dimensional positions within the growing area taken at one or more known times.

Abstract: A mobile platform includes at least one first sensor mounted at a first position on the mobile platform and at least one second sensor mounted at a second position on the mobile platform, where the first position is offset from the second position. The at least one first sensor is configured to capture first data measurements of plants in the growing area. The at least one second sensor is configured to capture second data measurements of the plants in the growing area. Each of the first and second data measurements is associated with a three-dimensional position within the growing area and a time. The first and second sensors are configured to generate at least one common type of data measurement such that at least some of the first and second data measurements represent stereo-spatio-temporal data measurements of the plants in the growing area.

Abstract: An apparatus includes at least one processor configured to obtain multiple spatiotemporal population projection models. Different spatiotemporal population projection models are associated with different pests, diseases, or biocontrol agents. Each spatiotemporal population projection model defines how the associated pest, disease, or biocontrol agent spreads and contracts in a growing area over time. The at least one processor is also configured to receive information associated with an actual presence of a specific pest, disease, or biocontrol agent at one or more monitored spatial locations in the growing area. Different monitored spatial locations in the growing area are associated with different plants. The at least one processor is further configured to project a future presence of the specific pest, disease, or biocontrol agent in the growing area using the spatiotemporal population projection model associated with the specific pest, disease, or biocontrol agent.

Abstract: A method includes obtaining data measurements associated with plants in at least one growing area. The plants have a common genotype and are grown under different growing or environmental conditions in the at least one growing area. The data measurements are associated with one or more characteristics of the plants and multiple characteristics of the growing or environmental conditions. The method also includes processing at least some of the data measurements to identify one or more of the growing or environmental conditions associated with at least one desired characteristic being expressed in the plants being grown. The method further includes outputting an identification of the one or more growing or environmental conditions identified as achieving a specific genotype or phenotype trait for the plants. The specific genotype or phenotype trait is associated with the at least one desired characteristic.

Abstract: An apparatus includes at least one processor configured to obtain multiple spatiotemporal population projection models. Different spatiotemporal population projection models are associated with different pests, diseases, or biocontrol agents in a growing area. Each spatiotemporal population projection model defines how the associated pest, disease, or biocontrol agent spreads and contracts in the growing area over time. The at least one processor is also configured to receive information associated with an actual presence of a specific pest, disease, or biocontrol agent at one or more locations in the growing area. Different locations in the growing area are associated with different plants. The at least one processor is further configured to project a future presence of the specific pest, disease, or biocontrol agent in the growing area using the spatiotemporal population projection model associated with the specific pest, disease, or biocontrol agent.

Abstract: A method includes receiving manual input from a human operator, where the manual input includes human observation data measurements associated with plants in a growing area. The method also includes, for each human observation data measurement, identifying a height of a movable portion of a mobile platform on which the human operator rides, identifying a location of the mobile platform in the growing area, identifying a time at which the human observation data measurement is received, and associating the human observation data measurement with a three-dimensional position within the growing area and the time in order to generate a spatio-temporal data measurement. The three-dimensional position includes the height and the location. The method further includes storing, processing, or transmitting at least some of the spatio-temporal data measurements or one or more results based on at least some of the spatio-temporal data measurements.

Abstract: A method and system for reinforcement learning can include an actor-critic framework comprising an actor and a critic, the actor comprising an actor network and the critic comprising a critic network; and a controller comprising a neural network embedded in the actor-critic framework and which can be tuned according to reinforcement learning based tuning including anti-windup tuning.

Abstract: An apparatus includes at least one processor configured to obtain stereo-spatio-temporal data measurements of plants in a growing area. The stereo-spatio-temporal data measurements include (i) first spatio-temporal data measurements of the plants in the growing area and (ii) second spatio-temporal data measurements of the plants in the growing area. The at least one processor is also configured to analyze the stereo-spatio-temporal data measurements to identify one or more actual or potential problems associated with one or more of the plants. The at least one processor is further configured to generate a graphical user interface identifying at least one of the one or more actual or potential problems with the one or more plants. The first and second spatio-temporal data measurements of each stereo-spatio-temporal data measurement are associated with at least one common plant characteristic and different three-dimensional positions within the growing area taken at one or more known times.

Abstract: A mobile platform includes at least one first sensor mounted at a first position on the mobile platform and at least one second sensor mounted at a second position on the mobile platform, where the first position is offset from the second position. The at least one first sensor is configured to capture first data measurements of plants in the growing area. The at least one second sensor is configured to capture second data measurements of the plants in the growing area. Each of the first and second data measurements is associated with a three-dimensional position within the growing area and a time. The first and second sensors are configured to generate at least one common type of data measurement such that at least some of the first and second data measurements represent stereo-spatio-temporal data measurements of the plants in the growing area.

Abstract: A method includes receiving manual input from a human operator, where the manual input includes human observation data measurements associated with plants in a growing area. The method also includes, for each human observation data measurement, identifying a height of a movable portion of a mobile platform on which the human operator rides, identifying a location of the mobile platform in the growing area, identifying a time at which the human observation data measurement is received, and associating the human observation data measurement with a three-dimensional position within the growing area and the time in order to generate a spatio-temporal data measurement. The three-dimensional position includes the height and the location. The method further includes storing, processing, or transmitting at least some of the spatio-temporal data measurements or one or more results based on at least some of the spatio-temporal data measurements.

Abstract: A method and system for reinforcement learning can involve applying a finite-difference approach to a controller, and tuning the controller in response to applying the finite-difference approach by taking a state as an entirety of a closed-loop step response. The disclosed finite-different approach is based on a random search to tuning the controller, which operates on the entire closed-loop step-response of the system and iteratively improves the gains towards a desired closed-loop response. This allows for prescribing stability requirement into the reward function without any modeling procedures.

Abstract: A method includes obtaining data measurements associated with plants in at least one growing area. The plants have a common genotype and are grown under different growing or environmental conditions in the at least one growing area. The data measurements are associated with one or more characteristics of the plants and multiple characteristics of the growing or environmental conditions. The method also includes processing at least some of the data measurements to identify one or more of the growing or environmental conditions associated with at least one desired characteristic being expressed in the plants being grown. The method further includes outputting an identification of the one or more growing or environmental conditions identified as achieving a specific genotype or phenotype trait for the plants. The specific genotype or phenotype trait is associated with the at least one desired characteristic.

Abstract: A method includes obtaining data measurements associated with plants in at least one growing area. The data measurements are associated with one or more characteristics of the plants and one or more characteristics of the at least one growing area. The method also includes processing at least some of the data measurements to identify one or more anomalous plant production issues associated with the plants. At least one anomalous plant production issue is associated with uneven production by the plants in at least part of the at least one growing area. The method further includes generating at least one visualization presenting at least some of the data measurements or processed versions of at least some of the data measurements in a spatial manner that corresponds to the at least one growing area. The at least one visualization identifies information associated with the one or more anomalous plant production issues.

Abstract: A method includes obtaining data measurements associated with plants in at least one growing area. The data measurements are associated with a characteristic of the plants or the at least one growing area that varies based on a time of day. The method also includes identifying a baseline that indicates how the characteristic varies based on the time of day. The method further includes processing at least some of the data measurements based on the baseline to at least partially reduce an effect of the time of day and at least partially remove time of day variations from at least some of the data measurements. In addition, the method includes using at least some of the processed data measurements to perform at least one function related to the plants or the at least one growing area.

Abstract: An apparatus includes at least one processor configured to obtain multiple spatiotemporal population projection models. Different spatiotemporal population projection models are associated with different pests, diseases, or biocontrol agents in a growing area. Each spatiotemporal population projection model defines how the associated pest, disease, or biocontrol agent spreads and contracts in the growing area over time. The at least one processor is also configured to receive information associated with an actual presence of a specific pest, disease, or biocontrol agent at one or more locations in the growing area. Different locations in the growing area are associated with different plants. The at least one processor is further configured to project a future presence of the specific pest, disease, or biocontrol agent in the growing area using the spatiotemporal population projection model associated with the specific pest, disease, or biocontrol agent.

Abstract: A method and system for reinforcement learning can involve applying a finite-difference approach to a controller, and tuning the controller in response to applying the finite-difference approach by taking a state as an entirety of a closed-loop step response. The disclosed finite-different approach is based on a random search to tuning the controller, which operates on the entire closed-loop step-response of the system and iteratively improves the gains towards a desired closed-loop response. This allows for prescribing stability requirement into the reward function without any modeling procedures.

Abstract: A method and system for reinforcement learning can include an actor-critic framework comprising an actor and a critic, the actor comprising an actor network and the critic comprising a critic network; and a controller comprising a neural network embedded in the actor-critic framework and which can be tuned according to reinforcement learning based tuning including anti-windup tuning.