Share this article:     

Coatings formulators and manufacturers are facing pressures, old and new in order to stay profitable in today’s business environment. Like many other enterprises, coatings companies are turning to Artificial Intelligence (AI) resources to meet these challenges. How can AI be used to develop coatings formulations? Can AI help overcome long-standing cost and performance challenges? Yes, by using AI coatings producers can meet the new challenges posed by sustainability initiatives and changing regulatory pressures.

According to the IBM website: “Artificial intelligence (AI) is technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy.” ¹ Let’s see what an AI agent says about AI in coatings formulation. I asked ChatGPT to provide a summary of the topic, “AI in Coatings Formulation” The popular AI agents reply was, “AI in Coatings Formulation refers to using machine learning and data-driven models to design, optimize, and troubleshoot paint and coating systems faster and more accurately than traditional trial-and-error methods.” ChatGPT offered to prepare a one slide graphic explaining AI in coatings formulation which is shown as the header image for this article.

Applications of AI in Coatings – Present and Future

For an expert’s insights and opinions, I spoke with Dr. Erik Sapper, Associate Professor of Chemistry at California State Polytechnic University in San Luis Obispo. Erik is a leading academic researcher on the topic of using AI and Machine Learning (ML) to enable better polymer synthesis and coatings formulations. We discussed the topic of AI in coatings formulation and addressed areas such as current applications, what he thinks the future will bring, and how the use of AI in the coatings industry will affect the workforce. Dr. Sapper teaches in Cal Poly SLO’s undergraduate and graduate programs in polymers and coatings.

The Current State of AI Implementation in Coatings

First Dr. Sapper and I talked about how AI is most commonly being utilized in coatings formulations today. “Where these tools are being used most effectively is as a lab assistant.” In his experience “the biggest and easiest gains are going to be in adjusting well-defined data sets to solve formulation problems using limited existing data.” This illustrates one of the common advantages to AI utilization cited by numerous sources, using AI as a tool to increase resource efficiency and reduce product development timelines.

AI as a data gathering and analysis tool, AI as a predictive tool

When asked about what’s next in the implementation of AI in the coatings industry, Dr. Sapper had this to say. “There will be a lot of options for data ingestion, taking existing data, i.e. spreadsheet data, and use data in different formats.” Using technologies such as Generative AI and Large Language Models (LLM) “data wrangling will be sped up, combining and collapsing disparate data sets into a singular data set for training models”. These models are then used to make predictions of formulation properties ranging from rheological and optical properties to film hardness and corrosion resistance.

ULTRUS™ Product Stewardship enables material selection, compliance, sustainability and chemical management throughout the product life cycle to meet regulatory requirements and consumer demands. Learn more here!

Cautions – AI in Coatings Formulation

Like any formulation technique based on predictions derived from data, data quality is all important. Small datasets can yield reliable predictions as long as those datasets are well-defined and documented. Specialized AI agents tailored to applications in chemistry or coatings can be very helpful as opposed to big data applications with very large, generalized datasets. Dr. Sapper remarked that development of new AI agents is taking place in both private industry and at universities like Cal Poly. Universities are taking the lead in areas like AI assisted polymer development while software developers are mainly working on generalized agents with broader market appeal. To quote Erik Sapper,” University researchers aren’t trying to sell anything, they can focus on the fundamentals.”

DIY AI Paint Formulation

A custom made of formulation targeted commercial AI agent isn’t required to get your feet wet in AI paint formulation. Try using general generative AI programs to guide your coatings formulation development efforts. Just as you might use a Google search to find raw materials or suppliers, you can use Googles AI Mode or ChatGPT to generate starting formulation ideas to build upon. A generative AI program is only as good as the input query and the data accessible to it. The user has no control over the dataset used by a general utility AI and therefore must tailor the input to get the most useful output.

Different AI agents might give completely different answers to the same query. As an example, let’s ask the Google AI Mode search and ChatGPT the same query. Using the query,“ “Formulate a mid-market industrial coating” I received these initial outputs.

Google AI Mode

“To formulate a midmarket industrial coating, you must balance cost and performance. This formula uses a styrenated acrylic emulsion to offer good chemical resistance and gloss at an affordable price point.

Base Formula Overview

Step-by-Step Manufacturing Process

1. The Grind Phase (High-Speed Dispersion)

• Add water, dispersant, and half of the defoamer to the tank.
• Turn on the high-speed disperser to a low speed.
• Slowly add titanium dioxide and calcium carbonate pigments.
• Increase blade speed to maximize shear.
• Mix until the grind reaches <25 microns on a Hegman gauge.

2. The Letdown Phase (Blending)

• Reduce the mixer speed significantly.
• Pump in the styrenated acrylic resin slowly.
• Premix the Texanol coalescent with a small amount of water.
• Add the Texanol blend slowly to avoid shocking the resin.

3. Final Adjustments

• Add the remaining defoamer and the flash rust inhibitor.
• Introduce the HEUR thickener slowly to reach target viscosity.
• Check and adjust the pH to a range of 8.5 to 9.0 using ammonia.

Key Performance Specifications

• Solids by Volume: ~38%
• VOC Content: <150 g/L
• Dry-to-Touch Time: 30 minutes
• Gloss Level: Semi-gloss (50-60 units at 60°)

To tailor this coating to your exact production environment, please let me know:

• What is your target substrate? (e.g., bare steel, aluminum, wood)
• What is the application method? (e.g., airless spray, roller, dip)
• What environmental exposure must it survive? (e.g., indoor UV, high humidity)”

ChatGPT

“Mid‑Market Industrial Coating Formulation Framework

Objective
Develop a cost-effective, durable industrial coating targeted at the mid-market segment with balanced:

• Corrosion resistance
• Chemical resistance
• Exterior durability
• Application robustness
• Competitive raw material cost
• Broad OEM / fabrication applicability

This example focuses on a solventborne direct-to-metal (DTM) acrylic-polyester industrial enamel suitable for:

• Agricultural equipment
• General metal fabrication
• Light industrial machinery
• Structural steel
• Trailers and utility equipment

Target Product Profile

Recommended Resin Platform

Primary Binder Choice

Acrylic Modified Polyester

Advantages:

• Good balance of cost and performance
• Better gloss retention than alkyds
• Lower cost than fluoropolymers or 2K urethanes
• Good pigment wetting
• Broad formulating latitude

Recommended resin characteristics:

Optional upgrades:

• Silicone modification for weatherability
• IBOMA incorporation for hardness + UV resistance
• Carbamate functionality for crosslink density

Crosslinker Selection

Melamine Crosslinker

Recommended for bake systems:

• Good hardness
• Good chemical resistance
• Low cost
• Strong appearance properties

Typical loading:

• 20–35% on binder solids

Alternative:

Polyisocyanate

For higher performance:

• Improved chemical resistance
• Better flexibility
• Higher corrosion resistance

Tradeoff:

• Higher cost
• Pot-life management
• More complex application

Example Starting Formulation

Part A — Pigment Grind

Grind Target

• Hegman: 6+
• Temperature: < 50°C

Part B — Letdown

Pigment Strategy

Cost-Optimized Pigment Package

Titanium Dioxide

Use:

• Chloride process rutile grades
• Durable exterior grades

Optimization strategy:

• Partial replacement with extender pigments
• Maintain PVC below CPVC

Extenders

Recommended:

• Barytes
• Nepheline syenite
• Talc
• Wollastonite

Benefits:

• Reduced TiO2 demand
• Improved film build
• Better sanding and application
• Improved corrosion resistance (wollastonite)

Corrosion Protection Package

Primary Inhibitor

Zinc Phosphate

Advantages:

• Regulatory acceptance
• Good DTM performance
• Lower hazard profile than chromates

Typical loading:

• 4–10% total formula

Optional synergists:

• Calcium ion exchange pigments
• Silicate inhibitors
• Modified phosphosilicates

Additive Package

Solvent Selection

Balanced Solvent Blend

Typical blend:

• Aromatic 100
• Xylene
• n-Butanol
• PM acetate

Goals:

• Proper flash profile
• Good atomization
• Flow and leveling
• Controlled dry time

Performance Optimization Levers

Improve Corrosion Resistance

• Increase zinc phosphate
• Lower PVC
• Add lamellar extenders
• Increase crosslink density
• Improve substrate wetting

Improve Exterior Durability

• Add HALS + UVA package
• Use acrylic-rich backbone
• Introduce IBOMA monomer
• Reduce aromatic content

Improve Cost Position

• Increase extender ratio
• Optimize TiO2 spacing
• Reduce specialty additive loadings
• Use hybrid resin architecture

Improve Application Window

• Broaden solvent evaporation profile
• Add associative rheology modifier
• Reduce high-shear viscosity
• Improve defoaming package

Typical Testing Program

Physical Testing

Corrosion Testing

Weathering

Recommended Mid‑Market Positioning

Competitive Space

This type of coating competes against:

• Alkyd enamels
• Economy urethanes
• Basic epoxies
• Imported industrial enamels

Value Proposition

• Better durability than alkyds
• Lower cost than premium urethanes
• Easier application than 2K systems
• Good appearance retention
• Broad substrate compatibility

Potential Upgrade Paths

Higher Performance Version

Move toward:

• 2K acrylic urethane
• Epoxy primer + urethane topcoat
• Waterborne acrylic urethane

Sustainability Version

Options:

• Waterborne backbone
• Bio-based monomers
• Exempt solvent strategy
• Lower bake temperature cure
• Heavy metal free pigments

Recommended Development Sequence

1. Define target cost per gallon
2. Define target corrosion performance
3. Select resin backbone
4. Optimize PVC
5. Build inhibitor package
6. Optimize application viscosity
7. Validate accelerated weathering
8. Benchmark against competitive products
9. Pilot production scale-up
10. Customer field validation

Example Mid‑Market Positioning Matrix

Key Formulation Philosophy

“A successful mid‑market industrial coating is usually not optimized for one extreme property. The winning balance is:

• Sufficient corrosion resistance
• Good enough weatherability
• Reliable application
• Competitive raw material cost
• Broad manufacturing tolerance
• Consistent supply chain availability

For many industrial OEMs and fabricators, robustness and cost stability are as important as peak technical performance.”

The same query posed to two AI agents resulted in very different outputs. Google AI modes response was shorter and resembled a recommended starting point formula commonly published by raw material suppliers. The ChatGPT output was a lengthy and detailed exploration of both technical and commercial aspects of the query. Each agent took its own approach to formulating a mid-market industrial coating. Google AI recommended a waterborne DTM formulation based on a styrenated acrylic latex while ChatGPT suggested a 2K approach but also gave recommendations for bake coatings. Follow-up queries can be used to refine the response.

A general AI agent won’t be as useful for optimizing a formulation based on previous trials as an application tailored to the task but can provide a starting point. A more specific initial query will help narrow down the response. For example, instead of asking for a “mid-market industrial coating formulation” you could request a “waterborne, medium cost and performance coating formulation for industrial applications”. This more specific query should result in responses with a narrower range of formulation and raw material parameters than a more general query.

Coupling AI Driven Formulation with Laboratory Automation

Failure to properly vet AI generated formulations is a pitfall that must be avoided no matter how good the “pencil formula” looks. In all cases formulations developed using AI agents still need to be physically made in the laboratory, usually by a human chemist or lab technician. This said, the coatings industry and its’ raw material suppliers are increasingly using laboratory automation coupled with AI formulation techniques to further shorten product development timelines. Laboratory automation can be defined as the integration of automated technologies into the laboratory to enable improved processes. Leading coatings industry suppliers including Clariant and Covestro are using laboratory automation to speed up processes such as pigment dispersion, color matching, and polymer/crosslinker optimization studies.

Fear of AI

Seventy three percent of Americans polled in 2025 believe that AI will reduce the total number of jobs over the next decade ². Clearly there are real concerns that the wide-spread implementation of AI and its effects on the workforce. While these worries are real, they are mostly about changes that haven’t occurred yet. When asked if AI is going to completely replace the coatings formulator Erik Sapper replied “Definitely not, 2026 is the year of adoption and implementation” of AI in coatings formulation. Even when AI formulation is coupled with the automated laboratory, human expertise will still be needed.

¹ https://www.ibm.com/think/topics/artificial-intelligence

² https://news.gallup.com/poll/694688/trust-businesses-improves-slightly.aspx?campaign_id=9&emc=edit_nn_20260102&instance_id=168721&nl=the-morning&regi_id=62380795&segment_id=213021&user_id=b9688bed2adbe006894c3e74d4c245f5

The views, opinions and technical analyses presented here are those of the author or advertiser, and are not necessarily those of ULProspector.com or UL Solutions. The appearance of this content in the UL Prospector Knowledge Center does not constitute an endorsement by UL Solutions or its affiliates.

UL Solutions does not make any representations or warranties with respect to the accuracy, applicability, fitness or completeness of the content. UL Solutions does not warrant the performance, effectiveness or applicability of sites listed or linked to in any content.

Share this article:     