Machine-Learning Chemical Formulation Seeds for Faster R&D

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Chemical and pharmaceutical industries face challenges in shortening product development lifecycles due to trial and error methods, stringent regulatory compliance, and a shortage of skilled manpower, exacerbated by an aging workforce, which hinders the generation of new product formulations.

Innovation Solution

A chemical product formulation system using machine learning (ML) to automatically generate seed formulae from historic experiments data, incorporating data preprocessing, supervised ML models, and analytical rules to ensure compliance with regulatory requirements, enabling efficient synthesis of chemical products.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If trial and error methods are used for developing new chemical products, then product formulations can be generated through experimentation, but the product development lifecycle is extended and time consumption increases

Engineering Contradiction:
Improveproduct formulation validityVSAvoidproduct development lifecycle
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary analysis by training machine learning models on historical experiment data before actual product development. This pre-training phase extracts analytical rules and builds predictive models that guide subsequent formulation development, eliminating the need for extensive trial-and-error experimentation and significantly shortening the development lifecycle while maintaining formulation validity

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates a virtual copy of the chemical formulation process through machine learning models that replicate the behavior and outcomes of physical experiments. These digital twins allow researchers to simulate and predict formulation results computationally, reducing the need for repeated physical trials and accelerating product development without sacrificing reliability

Inventive Principle:
Principle #26Copying

2Productivity

If more skilled manpower is deployed in R&D divisions, then new product formulations can be generated more effectively, but the shortage of skilled manpower and aging workforce constraints prevent this

Engineering Contradiction:
Improveformulation generation efficiencyVSAvoidworkforce structure complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The machine learning system serves itself by automatically training on historical data, generating analytical rules, and producing formulation recommendations without requiring extensive manual intervention. The system autonomously learns from past experiments and applies this knowledge to new formulation challenges, compensating for the shortage of skilled manpower while maintaining high productivity

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system replaces the mechanical dependency on skilled human researchers with an automated machine learning-based formulation system. The ML models perform the intellectual work of analyzing historical data, identifying patterns, and generating formulation recommendations, thereby decoupling productivity from the availability of skilled manpower and addressing the aging workforce challenge

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

3Reliability

If stringent compliance policies are enforced, then regulatory requirements are met and product quality is ensured, but the formulation development process becomes more complex and time-consuming

Engineering Contradiction:
Improveregulatory complianceVSAvoidformulation process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system incorporates compliance checking into the preliminary model training phase by including regulatory requirements as constraints in the objective function. This ensures that formulations generated by the ML system are pre-screened for compliance with quality standards and regulatory policies, eliminating the need for separate compliance verification steps and reducing overall process complexity

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The machine learning system performs multiple functions simultaneously: it optimizes formulation performance, ensures regulatory compliance, and generates interpretable analytical rules all within a single unified framework. The multi-objective optimization approach integrates compliance constraints directly into the formulation generation process, maintaining reliability while avoiding the need for separate compliance checking procedures

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

Data Source

PatentEP3896699B1Using machine learning for generating chemical product formulations
Publication Date: 2025.09.10 ACCENTURE GLOBAL SOLUTIONS LTD
  • EP3896699B1 patent drawingFigure 1
  • EP3896699B1 patent drawingFigure 2
  • EP3896699B1 patent drawingFigure 3

AI summary

A chemical product formulation system automatically generates seed formulae from historic experiments data for the synthesis of a chemical product. Independent and dependent features are identified from the historic experiments data and feature importance scores are calculated using a supervised machine learning (ML) model. The feature importance scores are used to build data structures from which analytical rules are extracted. The analytical rules are further processed to derive the seed formulae which are user-editable. The intermediate formulae generated via user edits of the seed formulae are further validated and approved in order to be used as the final formulae which are employed for the synthesis of the chemical product.