ML-Based Bale Property Prediction Using Chamber Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is an ongoing demand to improve the quality and quantity of harvested crop or plant material, particularly in baled form, where existing technologies lack the ability to accurately predict properties of baled plant material.
Innovation Solution
A computer-implemented method using a machine-learning algorithm to process sensor data from a baling chamber, enabling the prediction of properties such as the ratio of stem to leaves in baled plant material.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional baling machinery is used without advanced sensing and prediction systems, then the system remains simple and cost-effective, but the ability to accurately predict properties of baled plant material is lacking
Solution Approach 1:
The system performs preliminary sensing and prediction during the baling process itself, capturing sensor data from plant material before it is fully baled and predicting properties in advance. This allows for real-time monitoring and potential adjustments without requiring complex post-baling analysis systems.
Solution Approach 2:
A machine learning model serves as an intermediary between raw sensor data and property prediction outcomes. The model processes sensor readings from the baling chamber and translates them into meaningful predictions about baled plant material properties, bridging the gap between simple sensing and accurate prediction.
2Loss of information
If sensor data is collected and processed in real-time during baling, then granular information about plant material content is obtained, but the processing time and computational requirements increase
Solution Approach 1:
Sensor data is collected and preliminary processing is performed during the baling operation itself, rather than after completion. This real-time approach captures granular information about different portions of plant material as they are baled, enabling detailed analysis without requiring separate post-processing sessions.
Solution Approach 2:
The baled plant material is analyzed in segmented portions, with sensor data collected for each portion as it enters the baling chamber. This segmentation allows for granular information about specific sections of the bale while enabling parallel or sequential processing that manages computational load effectively.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
A mechanism for predicting one or more values for a property of a bale. A camera, positioned in a baling chamber of baling machinery, generates sensor data. The sensor data is processed using a machine-learning method to predict the one or more values.