ML-Controlled Enclosed Crop Growing Device
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional machine learning models for crop growth in enclosed systems face challenges in optimizing yield and quality for diverse crops at different growth stages under shared environmental conditions, requiring manual and uniform control methods that limit crop diversity and continuous harvesting.
Innovation Solution
A computer-implemented method using a machine learning model that processes images of crops within an enclosed growing device to automatically adjust environmental controls by extracting features, annotating crop types, growth stages, and health, and predicting optimal parameters for irrigation, temperature, and light to maximize yield and quality across multiple crop types and stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual and uniform control methods are used for environmental conditions, then the system is simple to operate, but crop diversity and continuous harvesting are limited
Solution Approach 1:
The system uses machine learning models to automatically analyze images of crops and determine optimal environmental parameters without human intervention. The ML model self-services by taking images as input and directly outputting control parameters for environmental components, eliminating the need for manual uniform control while enabling customized conditions for diverse crops at different growth stages.
Solution Approach 2:
The system dynamically changes environmental parameters (temperature, humidity, light, irrigation) based on ML model predictions that are customized for each crop type and growth stage. Instead of uniform parameters, the system adjusts parameters individually for each crop location, enabling crop diversity and continuous harvesting while automating the complexity of parameter management.
2Productivity
If uniform environmental control is applied to all crops, then the control system is simple, but yield optimization for diverse crops at different growth stages is limited
Solution Approach 1:
The system segments the enclosed growing space into multiple zones, each with customized environmental parameters determined by the ML model. Each crop or group of crops at similar growth stages receives tailored environmental conditions, allowing yield optimization for diverse crops while the ML model automates the complexity of managing multiple segmented zones.
Solution Approach 2:
The environmental control system becomes dynamic, with parameters continuously adjusted based on real-time image analysis and ML model predictions. The system adapts environmental conditions as crops progress through different growth stages, enabling yield optimization for diverse crops while the automation reduces the operational complexity of managing dynamic conditions.
3Adaptability or versatility
If automated image-based control is implemented, then crop diversity and yield optimization are improved, but system complexity and processing requirements increase
Solution Approach 1:
The system replaces manual mechanical control with an automated information-processing system based on machine learning. Images are captured and processed by ML models that automatically generate environmental control parameters, substituting the simplicity of manual control with the intelligence of automated systems that handle the complexity of customized environmental management for diverse crops.
Data Source
AI summary
A computer implemented method of automatically controlling an enclosed crop growing device, comprising: feeding an image of the enclosed crop growing device into a machine learning model, wherein the image simultaneously depicts a plurality of crops of a plurality of different types at a plurality of different growth stages arranged at a plurality of predefined crop growing locations within the enclosed crop growing device, obtaining as an outcome of the ML model, a plurality of parameters for setting a plurality of environmental control components that control an environment of the enclosed crop growing device, and automatically adjusting the plurality of environmental control components according to the plurality of parameters.