Formulation Development with Artificial Intelligence

Compound development processes require significant time and numerous iterations to find the optimal formulation. 

Uncountable enables R&D labs to cut development time and reduce testing iterations

Artificial intelligence algorithms suggest the best recipes to run, avoiding the tedious, manual tweaking of individual ingredients. Complicated development processes that used to involve tens or hundreds of experiments can now be conducted in half the time utilizing advanced machine learning models.

 

HOW IT WORKS

list-1.png

step 1: ingest previous test data

Uncountable's software incorporates multiple data sources to understand every piece of the development process, including property objectives, testing requirements, list of available ingredients, past formulations and test history. 

Past experimental data can be ingested via several methods, including via database dumps, disparate spreadsheets, or other record systems. 


step 2: model material space

Uncountable's machine learning algorithms build out a model of the material property performance "space." The use of artificial intelligence models allows the system to understand the effect that altering each subset of ingredients will have on the performance of the formulation.

Models from traditional DOE software that typically rely on linear regressions make oversimplifications and fail to capture the complexity of the development process.


map-location.png

step 3: suggest new recipes

Artificial intelligence software delivers multiple new compound formulations to the user for further testing. These suggestions are the result of a high dimensional optimization analysis, yielding optimized testing sequences and reduced trial cycles en route to the ideal material recipe.  

This obviates the need for factorial testing where every combination of values is tried. Uncountable's approach is adaptive, enabling significantly more learnings per round of experiments while also moving closer to the project property goals.

The machine learning engine is continuously improved by feeding further trial information into its base of test history, meaning a research team gets better and better at finding breakthrough materials. 


How our technology differs from a traditional design of experiments (DOE)

Traditional DOE's

Tools: 

  • Excel
  • DOE software like JMP & Mini-Tab

Modeling Techniques: 

  • Linear regressions
  • Fixed parametric models 

Formulation Testing Procedure: 

  • Orthogonal designs - vary few parameters at a time
  • Manual designs - somewhat arbitrary and manual proposal of new formulations 

 

Uncountable's Approach

Tools: 

  • Custom software designed for the material development experimental cycle

Modeling Techniques: 

  • Machine learning models
  • Flexible, non-parametric Bayesian models

Formulation Testing Procedure: 

  • New formulations chosen via artificial intelligence after simulating thousands of combinations
  • Run experiments that are most likely to be successful while also providing more information to improve the model.

Download Our Case Study

Fill out the form below and check out our case study with Fortune 1000 automotive supplier Cooper Standard.  

Name *
Name

Frequently Asked Questions

How does Uncountable deliver suggestions for experiments?

We'll provide you everything you need to view and interact with our models. This starts with our cloud-based software solution that displays visualizations and reports customized to the needs of your company. Our platform can be used to record experimental results, visualize material performance, and create new projects with specific goals and constraints.

 

How is experimental data stored, and how do I know it's secure?

We understand that formulation records are at the core of our customers' businesses. We take every precaution to ensure that these essential records are securely stored in the cloud. Uncountable hosts all data in Amazon Web Services (AWS) and utilizes multiple security protocols to protect information, including industry-standard encryption practices as well as strict IP whitelisting and two-factor authentication for server access.

Uncountable can also work with anonymized or encrypted data, such that the cleartext names of individual ingredients are never transmitted or processed by Uncountable. In this way, no one outside your organization can ever see your true formulations. Please contact us at info@uncountable.com for more details.

 

How does intellectual property work?

Customers retain all IP in the data on Uncountable's platform including all formulations that are suggested. We never claim any ownership of new formulations, experimental data, or experimental results. 

 

How much does it cost and how does Uncountable charge?

Uncountable works on a subscription model.  We charge a monthly or quarterly fee for our modeling services and access to our web platform. There are no initial fees or costs for integrating data.

We strive to provide a high-touch solution, tailored to each customer. The subscription model allows us to continuously update and improve the efficacy of our formulation suggestion engine to address the specific needs of each of our customers.

Pricing depends on the scale of each deployment.  Please email us at info@uncountable.com to request a quote.

 

How do I try it?

Our initial engagements feature the customer selecting a few high-value development projects for Uncountable to collaborate on. We will work with you to establish project goals and to understand experimental procedures.  Uncountable's model will then be trained to suggest new rounds of experimental testing and support your scientists during the development process.


Want to learn more?

Contact our team to understand how Uncountable's technology could be used to expedite your formulation development process.

Name *
Name