Pet Food Recommendation Using Image Recognition and Machine Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing pet food recommendation systems are primarily filter-based, static, and lack personalization, failing to account for individual pet characteristics and changing circumstances over time, leading to suboptimal feeding decisions.

Innovation Solution

A pet food recommendation system utilizing image recognition and machine learning to analyze pet characteristics and preferences, providing personalized and proactive recommendations that adapt over time, considering factors like breed, size, activity level, and owner preferences.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If filter-based recommendation systems are used, then product selection is narrowed based on available characteristics, but the system fails to capture sufficient information to provide optimal solutions for individual pets

Engineering Contradiction:
Improvepersonalization capabilityVSAvoidinsufficient pet information capture
Core Design Contradiction:
Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent replaces traditional filter-based mechanical selection systems with an AI/ML-based recommendation system. The system uses machine learning models to process and analyze comprehensive pet information (images, video, health data, lifestyle factors) to generate personalized recommendations, substituting simple filtering with intelligent pattern recognition and prediction algorithms that adapt to individual pet needs over time

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

Solution Approach 2:

The patent introduces multiple intermediaries between the user and the recommendation output. These include AI/ML algorithms, data processing layers, and analytical models that mediate the transformation of raw pet data into actionable recommendations. The system uses intermediaries such as image recognition models, health data analyzers, and lifestyle factor processors to bridge the gap between collected information and personalized advice

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If static recommendation systems are used, then recommendations are provided at a given point in time, but they do not account for changes in the pet over time

Engineering Contradiction:
ImprovedynamismVSAvoidrelevance over time
Core Design Contradiction:
Adaptability or versatilityVSDuration of action of stationary object

Solution Approach 1:

The patent implements dynamic recommendation systems that continuously adapt to changing pet conditions. The system incorporates real-time data from pet wearables, periodic health checks, and ongoing lifestyle monitoring to update recommendations dynamically. The AI models continuously learn from new data to adjust their predictions, ensuring recommendations remain relevant as pets age, gain weight, change activity levels, or develop health conditions

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system incorporates multiple feedback loops where pet outcomes (weight changes, health metrics, behavior observations) are fed back into the recommendation algorithm. This feedback mechanism allows the system to learn what works for each individual pet and continuously refine its recommendations. Owners can also provide feedback on recommendation effectiveness, which further trains the AI models to improve future recommendations

Inventive Principle:
Principle #23Feedback

3Adaptability or versatility

If manufacturer product selectors are used, then recommendations are limited by the breadth of the portfolio of a single pet food manufacturer

Engineering Contradiction:
Improverecommendation breadthVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent creates a universal recommendation system that serves multiple functions: it analyzes pet characteristics, processes health data, evaluates lifestyle factors, generates personalized recommendations, and provides feeding guidance. The system is designed to work with diverse data sources and can recommend products from multiple manufacturers, making it a multi-functional platform that handles the entire pet nutrition decision-making process without being constrained to a single manufacturer's portfolio

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3909009B1Pet food recommendation devices and methods
Publication Date: 2025.09.24 SOCIETE DES PRODUITS NESTLE SA
  • EP3909009B1 patent drawingFigure 1
  • EP3909009B1 patent drawingFigure 2
  • EP3909009B1 patent drawingFigure 3

AI summary

A system or method is provided that includes receiving a pet image. The pet image may depict a pet and may be received from a user device. The system or method may further analyze the pet image with a pet image recognition model to determine one or more pet characteristics of the pet. In certain embodiments, analyzing the pet image may further include identifying one or more image characteristics of the pet image. The system or method may further analyze the pet characteristics with a pet food recommendation model to generate a pet food recommendation for the pet.