Harvester Subsystem Control with Qualified Real-Time Field Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for controlling work machines, such as combine harvesters, rely on a priori data for generating predictive models, which do not account for actual ground truth data, leading to inaccurate yield predictions and delayed model improvements, as actual yield data is only available after the harvesting operation is completed.
Innovation Solution
A system that generates a predictive model using both a priori geo-referenced vegetative index data and in situ field data collected during the operation, dynamically evaluating and updating the model's quality to ensure accurate control of the machine's subsystems in real-time, switching to alternative models or manual operation if the model quality is insufficient.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If a priori data is used to generate predictive models, then model generation is possible before operation, but model accuracy deteriorates due to lack of actual ground truth data
Solution Approach 1:
The system implements feedback by continuously comparing actual sensor data collected during harvesting operations with predictive model outputs. This feedback loop enables real-time model validation and adjustment, allowing the system to maintain accurate yield predictions while operating with time-efficient pre-generated models. The feedback mechanism resolves the contradiction by verifying model accuracy against ground truth data without requiring post-operation model regeneration.
Solution Approach 2:
The system performs preliminary model generation using a priori data before harvesting operations begin, enabling early model availability for operational planning. This preliminary action is complemented by real-time validation during operations, ensuring that the time-efficient pre-generated models do not sacrifice accuracy when compared against actual field data.
2Measurement precision
If predictive models are generated and evaluated in real-time during operations, then model accuracy improves through actual data, but system complexity increases
Solution Approach 1:
The control system is designed with multi-functionality to handle both real-time data collection and model evaluation without requiring separate dedicated systems. The same computing infrastructure that controls harvesting operations also performs predictive model generation, validation, and adjustment, reducing overall system complexity while maintaining real-time model accuracy through actual field data.
Solution Approach 2:
The system merges the predictive model evaluation function with the existing operational control system. Rather than adding a separate complex model validation system, the patent integrates model accuracy assessment into the normal operational data processing workflow, thereby improving model accuracy through real data while minimizing increases in system complexity.
3Reliability
If actual field data is collected and used to validate models during operations, then model quality improves, but data processing requirements increase
Solution Approach 1:
The system applies partial action by selectively validating predictive models only against critical field data parameters that most significantly impact model quality. Rather than processing and comparing all collected sensor data, the system identifies and processes only the essential subsets of data needed for effective model validation, thereby improving model quality while minimizing energy consumption for data processing.
Data Source
AI summary
A priori geo-referenced vegetative index data is obtained for a worksite, along with field data that is collected by a sensor on a work machine that is performing an operation at the worksite. A predictive model is generated, while the machine is performing the operation, based on the geo-referenced vegetative index data and the field data. A model quality metric is generated for the predictive model and is used to determine whether the predictive model is a qualified predicative model. If so, a control system controls a subsystem of the work machine, using the qualified predictive model, and a position of the work machine, to perform the operation.