Cultivar Growth Modeling for Predictive Pruning Decisions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Pruning in horticulture is a knowledge and resource-intensive task that requires precise decision-making to improve crop yields and quality, but existing methods lack efficient tools for predicting optimal pruning actions based on cultivar growth and environmental conditions.

Innovation Solution

A horticulture support platform that uses multi-dimensional modeling and image data analysis to generate predictive models of cultivar growth, recommending pruning actions by encoding cultivar data into strings of symbols based on an L-system or RNN/EDA approaches, and adjusting rules based on environmental context to optimize yield.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional pruning methods are used based on expert knowledge, then pruning quality can be maintained, but the process becomes resource-intensive and difficult to scale

Engineering Contradiction:
Improvepruning decision accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces manual expert knowledge and mechanical pruning processes with an automated computer vision system using deep learning models. The system captures images of cultivars, processes them through neural networks to identify growth patterns and predict optimal pruning actions, thereby substituting human expert judgment with algorithmic decision-making that reduces resource intensity while maintaining or improving precision

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables the cultivar itself to 'inform' the pruning decisions by automatically analyzing its own growth patterns through image processing. The deep learning model extracts features directly from images of the cultivar, allowing the system to self-determine optimal pruning actions without requiring external expert intervention for each decision

Inventive Principle:
Principle #25Self-service

2Productivity

If comprehensive image data and multi-dimensional modeling are used to improve pruning predictions, then yield accuracy improves, but computational resources and processing time increase

Engineering Contradiction:
Improvecrop yield prediction accuracyVSAvoidprocessing time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by continuously capturing and analyzing images of cultivars throughout their growth cycle. Deep learning models pre-process image data to extract growth patterns and predict future states, allowing the system to prepare pruning recommendations in advance rather than making rushed decisions at harvest time, thereby improving accuracy without proportionally increasing processing time

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the complex task of pruning decision-making into distinct processing stages: image capture, feature extraction through deep learning, growth pattern analysis, and prediction generation. This segmentation allows each component to be optimized independently, reducing overall processing time while maintaining comprehensive analysis for high yield prediction accuracy

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3617816B1Modeling and decision support for horticulture
Publication Date: 2022.11.09 ACCENTURE GLOBAL SOLUTIONS LTD
  • EP3617816B1 patent drawingFigure 1
  • EP3617816B1 patent drawingFigure 2
  • EP3617816B1 patent drawingFigure 3A~3B

AI summary

Implementations include providing a baseline multi-dimensional model of a cultivar, determining an encoding based on the baseline multi-dimensional model, and a target multi-dimensional model, the encoding defining a string of symbols, and being based on an alphabet and a set of rules, providing an expected multi-dimensional model based on the encoding, and a modified set of rules, the modified set of rules being based on the set of rules, the expected multi-dimensional model representing the cultivar after a period of time, selecting a set of actions by determining multiple predicted multi-dimensional models for each set of actions in a plurality of sets of actions, and, for each predicted multi-dimensional model, providing a predicted yield that can be used to determine impact with respect an expected yield, the set of actions being selected based on a respective impact, and providing the set of actions as output for executing on the cultivar.