머티리얼즈 인포매틱스(MI) 시장(2026-2036년)

Materials informatics (MI) has emerged as one of the most consequential transformations in industrial R&D since the digitalisation of design itself. Built on the convergence of materials science, data science, and artificial intelligence, MI applies machine learning, high-throughput computation, generative models, and large language models to compress the time and cost of discovering and optimising new materials. Industry practitioners now routinely report 50–70% reductions in the number of physical experiments required to develop a new material, with corresponding time-to-market acceleration measured in years rather than months. What once required decades of iterative trial-and-error can increasingly be completed in two-to-five-year programmes guided by data-driven workflows.

The market has moved from an early-adopter phase between 2014 and 2018, through a growth phase between 2019 and 2023, into the AI-boom acceleration phase that began in 2024 and now defines the industry. Three forces shape the 2026 landscape. First, foundation models, transformer architectures, generative diffusion models, and universal machine-learning interatomic potentials originally developed for language and vision have crossed over decisively into materials science. Second, big technology firms — Microsoft, Google DeepMind, Meta FAIR, IBM Research, and NVIDIA — have entered the field as direct competitors and infrastructure providers, reshaping competitive economics for the dedicated MI vendor category. Third, mega-funding rounds have arrived, with Lila Sciences alone raising approximately US$550 million cumulatively by Q1 2026 to build fully autonomous labs for life, chemical, and materials sciences.

Adoption is now mainstream. Virtually every major materials player has engaged with MI through external service providers, consortia membership, or in-house programmes. Executive-level mandates to demonstrate AI impact across the business have become as common as bottom-up scientist-led pilots. Sustainability-driven applications — catalysts for green hydrogen, sorbents for carbon capture, low-embodied-carbon cement, recyclable polymers, PFAS replacements, energy-transition battery and fuel-cell materials — represent the largest single application driver, accounting for an increasing share of programme spend through 2036.

The Global Materials Informatics Market 2026–2036 provides a comprehensive analysis of the materials informatics industry at its most transformative inflection point to date. Building on the methodology established in earlier editions and informed by primary interviews conducted with industry players through 2025–2026, this revised edition captures the structural reshaping of the field driven by foundation models, big-tech entry, and the commercialisation of self-driving laboratories. The report forecasts the market through 2036 with both a narrower external MI provider revenue segment and a broader total MI software and services market segment that captures big-tech cloud platform revenue, project-based services, and addressable in-house spend.

The report examines the technologies, business models, applications, and players that define the modern MI industry. New for 2026 is dedicated treatment of foundation models for materials science; the strategic implications of big-tech entry; the autonomous-laboratory revolution; the sharp bifurcation in the funding landscape between mega-rounds for integrated AI-and-experimentation platforms and headwinds facing first-generation MI SaaS; and the geopolitical context.

Report Contents

• Executive summary including 2026 industry state, AI-boom impact, and global market forecasts
• Introduction covering motivations, AI integration, and parallel informatics fields
• Technology analysis: algorithms, foundation models, generative AI, LLMs, agentic AI scientists
• Data infrastructure, databases (Materials Project, AFLOW, NOMAD, OMat24, GNoME), small-data strategies
• Computational materials science: DFT, ICME, universal MLIPs, quantum computing
• Autonomous experimentation and self-driving laboratories
• Twenty-eight application areas including alloys, drug discovery, batteries, catalysts, polymers, photovoltaics, carbon capture, PFAS replacement, critical minerals
• Industry analysis: strategic approaches, player categories, funding, SaaS economics, big-tech competition
• MI consortia and public-private initiatives globally
• Market forecasts with bull, base, and bear scenarios
• 53 company profiles
• Research methodology and references

Companies Profiled include Aionics, Albert Invent, Alchemy Cloud, Ansatz AI, Asahi Kasei, Atomic Tessellator, Citrine Informatics, Copernic Catalysts, Cynora, DeepVerse, Dunia Innovations, Elix Inc, Enthought, Exomatter GmbH, Exponential Technologies Ltd, FEHRMANN MaterialsX, fibclick, Genie TechBio, Google DeepMind GNoME, Hitachi High-Tech, IBM Research Materials, Innophore, Intellegens, Kebotix, Kyulux, LG AI Research, Lila Sciences, MaterialsZone, Matmerize Inc, Mat3ra, META, Microsoft, N-ERGY, Noble.AI, Novyte Materials and more......

Table of Contents

1 EXECUTIVE SUMMARY

• 1.1 What is Materials Informatics?
• 1.2 Materials Informatics: State of the Industry in 2026
• 1.3 Issues with Materials Science Data
• 1.4 Dealing with Little or Sparse Data
• 1.5 Key Technologies Driving Materials Informatics
• 1.6 Importance in Modern Materials Science and Engineering
• 1.7 Market Challenges and Restraints
• 1.8 Recent Industry Developments
• 1.9 The AI Boom and Its Impact on Materials Informatics
• 1.10 Foundation Models, Generative AI and Materials Discovery
• 1.11 Big Tech Entry into Materials Informatics
• 1.12 Market Players
• 1.13 Funding Landscape: Mega-Rounds and SaaS Headwinds
• 1.14 Future Markets Outlook and Opportunities
  • 1.14.1 Integration of AI and Robotics in Materials Labs
  • 1.14.2 Self-Driving Laboratories and Autonomous Science Platforms
  • 1.14.3 Quantum Machine Learning for Materials Discovery
  • 1.14.4 Blockchain for Materials Data Provenance
  • 1.14.5 Edge Computing in Materials Informatics
  • 1.14.6 Augmented and Virtual Reality in Materials Design
  • 1.14.7 Neuromorphic Computing for Materials Modeling
  • 1.14.8 Materials Informatics as a Service (MIaaS)
  • 1.14.9 Integration with Internet of Things (IoT)
  • 1.14.10 Green Technology and Circular Economy Applications
  • 1.14.11 Agentic AI Scientists
• 1.15 MI Roadmap
• 1.16 Economic Impact Analysis
  • 1.16.1 Cost Savings in Materials R&D
  • 1.16.2 Accelerated Time-to-Market for New Materials
  • 1.16.3 Job Creation and Skill Development
  • 1.16.4 Impact on Traditional Materials Industries
• 1.17 Sustainability and Environmental
  • 1.17.1 Role of Materials Informatics in Sustainable Development
  • 1.17.2 Reducing Environmental Impact of Materials Production
  • 1.17.3 Design for Recyclability and Circular Economy
  • 1.17.4 Bio-inspired Materials Discovery
  • 1.17.5 Materials for Energy Transition
• 1.18 Geopolitical Considerations: U.S., EU, China, Japan, Korea
• 1.19 Global Market Forecasts

2 INTRODUCTION

• 2.1 Advent of the Data Science Era
• 2.2 Background to the Emergence of MI
• 2.3 Motivation for Materials Informatics Development
  • 2.3.1 Accelerating Discovery
  • 2.3.2 Cost Reduction
  • 2.3.3 Addressing Global Challenges
  • 2.3.4 Maximizing Data Value
  • 2.3.5 Handling Complexity
  • 2.3.6 Enabling Targeted Design (Inverse Design)
  • 2.3.7 Improving Reproducibility
  • 2.3.8 Integrating Multidisciplinary Knowledge
  • 2.3.9 Supporting Sustainability
  • 2.3.10 Competitive Advantage
• 2.4 Integration of Artificial Intelligence (AI) into materials science and engineering
  • 2.4.1 AI Opportunities at Every Stage of Materials Design and Development
  • 2.4.2 The Transition from Predictive AI to Generative AI in Materials
  • 2.4.3 Physical AI: Models that Understand Physics and Chemistry
• 2.5 Problems with Materials Science Data
• 2.6 Algorithm Advancements
• 2.7 Materials Informatics Categories
• 2.8 Trend towards data-driven approaches in science and engineering
  • 2.8.1 Bioinformatics
  • 2.8.2 Cheminformatics
  • 2.8.3 Geoinformatics
  • 2.8.4 Health Informatics
  • 2.8.5 Environmental Informatics
  • 2.8.6 Astroinformatics
  • 2.8.7 Neuroinformatics
  • 2.8.8 Engineering Informatics
  • 2.8.9 Energy Informatics
  • 2.8.10 Quantum Informatics
• 2.9 Challenges
• 2.10 Advantages of Machine Learning
  • 2.10.1 Acceleration
  • 2.10.2 Scoping and Screening
  • 2.10.3 New Species and Relationships
  • 2.10.4 Closing the Loop on Traditional Synthetic Approaches
  • 2.10.5 High-Throughput Virtual Screening (HTVS)
• 2.11 Data Infrastructures for Chemistry and Materials Science
• 2.12 ELN/LIMS Software and Materials Informatics
• 2.13 Proving the Value of Materials Informatics: Case Studies

3 TECHNOLOGY ANALYSIS

• 3.1 Overview
  • 3.1.1 Inputs and Outputs of Materials Informatics Algorithms
  • 3.1.2 What is Needed for Materials Informatics?
• 3.2 Technology approaches
  • 3.2.1 Summary of Technology Approaches
  • 3.2.2 Uncertainty in Experimental Data
  • 3.2.3 Data Mining
  • 3.2.4 Machine Learning and AI
  • 3.2.5 High-Throughput Computation
  • 3.2.6 Data Infrastructure
  • 3.2.7 Visualization Tools
  • 3.2.8 Reinforcement Learning
  • 3.2.9 Natural Language Processing
  • 3.2.10 Automated Experimentation
  • 3.2.11 Workflow Management
  • 3.2.12 Quantum Computing
  • 3.2.13 QSAR and QSPR
  • 3.2.14 Automated feature selection
  • 3.2.15 Exploitation vs exploration
  • 3.2.16 Pure exploitation vs epsilon-greedy policies in materials informatics
  • 3.2.17 Active learning and MI: Choosing experiments to maximize improvement
• 3.3 MI Algorithms
  • 3.3.1 Overview of MI Algorithms
  • 3.3.2 Types of MI Algorithms
  • 3.3.3 Descriptors and Training a Model
  • 3.3.4 Supervised vs. Unsupervised Learning
  • 3.3.5 Automated Feature Selection
  • 3.3.6 Exploitation vs. Exploration; Active Learning
  • 3.3.7 Bayesian Optimization
  • 3.3.8 Genetic Algorithms
  • 3.3.9 Generative vs. Discriminative Algorithms
  • 3.3.10 Deep Learning and Neural Network Types
  • 3.3.11 Physics-Informed Neural Networks (PINNs)
  • 3.3.12 Graph Neural Networks (GNNs) for Materials
  • 3.3.13 Transformer Models and the AI Boom
  • 3.3.14 Foundation Models for Materials
    • 3.3.14.1 Definition and Architecture
    • 3.3.14.2 Foundation Models for Computational Data
    • 3.3.14.3 Foundation Models for Experimental Data
    • 3.3.14.4 Limitations: Data Availability and Compute Cost
  • 3.3.15 Generative Models for Inorganic Compounds
    • 3.3.15.1 Variational Autoencoders and GANs
    • 3.3.15.2 Diffusion Models for Crystal Generation
  • 3.3.16 Large Language Models (LLMs) and Materials R&D
    • 3.3.16.1 Capabilities of LLMs in Science
    • 3.3.16.2 LLM-Powered Material Data Mining
    • 3.3.16.3 Agentic LLMs and Autonomous Research
  • 3.3.17 AutoML: Democratizing Machine Learning
  • 3.3.18 Multi-Model Ensembles
  • 3.3.19 How to Work with Small Material Datasets
  • 3.3.20 Algorithmic Approaches in MI Are Diverse — Summary
• 3.4 Data infrastructure
  • 3.4.1 Overview
  • 3.4.2 Developments Targeted for Chemical and Materials Science
  • 3.4.3 ELN/LIMS Integration with MI Workflows
• 3.5 Databases and External Repositories
  • 3.5.1 Data Repositories — Organizations
  • 3.5.2 Leveraging Data Repositories
  • 3.5.3 The Materials Project, AFLOW, NOMAD, OQMD
  • 3.5.4 Meta's Open Materials 2024 (OMat24) Dataset
  • 3.5.5 GNoME Dataset and DeepMind's Contributions to the Materials Project
  • 3.5.6 Text Extraction and Analysis
  • 3.5.7 ChemDataExtractor V1.0 and V2.0
  • 3.5.8 LLMs Expand Material Data Mining Capabilities
• 3.6 Databases to Big Data
  • 3.6.1 Rapid data generation and collection
  • 3.6.2 Integrated use of materials databases
  • 3.6.3 Data reliability
• 3.7 Small Data Strategies in Materials Informatics
  • 3.7.1 Utilizing data correlations
  • 3.7.2 Selecting descriptors based on theory and experience
• 3.8 MI with Physical Experiments and Characterization
  • 3.8.1 High-Throughput Experimentation (HTE)
  • 3.8.2 In-situ and Operando Characterisation
  • 3.8.3 Advanced Imaging and Spectroscopy
  • 3.8.4 Why High-Throughput Screening in Materials is Tougher Than in Other Fields
• 3.9 Computational Materials Science
  • 3.9.1 Simulations for Chemistry and Materials Science R&D
  • 3.9.2 Density Functional Theory (DFT) — Quantum Mechanical Modeling
  • 3.9.3 Surrogate Models for Atomistic Simulation
  • 3.9.4 Universal ML Interatomic Potentials (CHGNet, MACE, M3GNet, MatterSim)
  • 3.9.5 Multiscale Modelling
  • 3.9.6 Integrated Computational Materials Engineering (ICME)
  • 3.9.7 ICME and the Role of Machine Learning
  • 3.9.8 QuesTek Innovations and ICME: From Service to SaaS
  • 3.9.9 Thermo-Calc, CompuTherm and the ICME Software Ecosystem
  • 3.9.10 Cloud-Based Simulation Platforms
  • 3.9.11 The Potential in Leveraging Quantum Computing
  • 3.9.12 Big Tech, Computational Materials Science and MI
• 3.10 Autonomous Experimentation and Self-Driving Labs
  • 3.10.1 The Vision: Fully Autonomous Labs
  • 3.10.2 The Chemputer
  • 3.10.3 Workflow Management for Laboratory Automation
  • 3.10.4 A-Lab (Lawrence Berkeley): Closed-Loop Synthesis Validation
  • 3.10.5 Lila Sciences AI Science Factory
  • 3.10.6 Dunia Innovations: Physics-Informed ML + Lab Automation
  • 3.10.7 Google DeepMind's Gemini-Powered Autonomous Lab
  • 3.10.8 Commercial Self-Driving Laboratories
  • 3.10.9 Mobile Autonomous Robots in Academia
  • 3.10.10 Retrosynthesis Through to Robot Execution
  • 3.10.11 Technology Pillars for Chemical Autonomy
• 3.11 Multi-modal Data Integration
• 3.12 Inverse Problems in Materials Characterization
• 3.13 Data-driven Experimental Design
• 3.14 Automated Data Analysis and Interpretation
• 3.15 Robotics and Automation in Materials Research
• 3.16 Digital Twins for Materials and Process Engineering

4 APPLICATIONS OF MATERIALS INFORMATICS

• 4.1 Alloy Design and Optimization
  • 4.1.1 High-Entropy Alloy Design
  • 4.1.2 Aluminum and titanium alloys
  • 4.1.3 Metallic glass alloys
  • 4.1.4 Nickel-base superalloys
  • 4.1.5 Steels for Extreme Environments
• 4.2 Drug Discovery and Development
  • 4.2.1 AI-Driven Drug Design
• 4.3 Intermetallics
• 4.4 Organometallics
• 4.5 Organic Electronics
  • 4.5.1 RFID
  • 4.5.2 OPV
  • 4.5.3 OLEDs
  • 4.5.4 Emerging Areas
• 4.6 Coatings and Paints
• 4.7 Catalysts
  • 4.7.1 Heterogeneous Catalysts
  • 4.7.2 Catalysts for Green Hydrogen Production
  • 4.7.3 Open Catalyst Project (Meta)
• 4.8 Ionic liquids
• 4.9 Battery Materials
  • 4.9.1 Lithium-ion batteries
  • 4.9.2 Solid-State Batteries
  • 4.9.3 Lithium-Sulfur and Beyond-Li Batteries
  • 4.9.4 Accelerated Battery Material Discovery
• 4.10 High-density Heat Storage Materials
• 4.11 Hydrogen-based Superconductors
• 4.12 Sorbents for Carbon Capture
• 4.13 Polymer Informatics
  • 4.13.1 Optimizing Additive Manufacturing Materials
  • 4.13.2 Sustainable Polymer Development
  • 4.13.3 Large Engineering Models for Polymer Processing
• 4.14 Rubber processing
• 4.15 Nanomaterials
  • 4.15.1 Nanofabrication
  • 4.15.2 Quantum Dots
  • 4.15.3 Other Nanomaterials
• 4.16 2D materials
• 4.17 Metamaterials
• 4.18 Lubricants
• 4.19 Thermoelectric Materials
• 4.20 Photovoltaics
  • 4.20.1 Light Absorbers and Solar Cells
  • 4.20.2 Perovskite Photovoltaics
  • 4.20.3 Tandem Cells
• 4.21 Metal-insulator transition compounds
• 4.22 Self-assembled monolayers
• 4.23 Construction Materials and Cement
• 4.24 Biomaterials
• 4.25 Materials for Quantum Technologies
• 4.26 Materials for Defence and Extreme Environments
• 4.27 PFAS Replacement Materials
• 4.28 Critical Minerals and Rare-Earth Substitution

5 INDUSTRY ANALYSIS

• 5.1 Materials Informatics: State of the Industry in 2026
• 5.2 Strategic Approaches to MI
  • 5.2.1 Materials Informatics Players
  • 5.2.2 SaaS Platforms
  • 5.2.3 Project-Based Consultancies
  • 5.2.4 In-house Development by Materials Corporates
  • 5.2.5 Big Tech Cloud Platforms
  • 5.2.6 Conclusions for End-Users
  • 5.2.7 Conclusions for External MI Companies
• 5.3 Player Analysis
  • 5.3.1 Materials Informatics Players — Overview
  • 5.3.2 Key Partners and Customers of Selected External Providers
  • 5.3.3 Partnerships with Engineering Simulation Software
  • 5.3.4 Funding Raised by Private Companies (I): In-House Development Drives Capital Requirements
  • 5.3.5 Funding Raised by Private Companies (II): The State of SaaS Business Models
  • 5.3.6 Pricing MI SaaS Platforms
    • 5.3.6.1 Risks for SaaS Business Models in MI
  • 5.3.7 Barriers to Profitability for MI SaaS Players
  • 5.3.8 Microsoft's Azure Quantum Elements: Competition for Dedicated MI Players
  • 5.3.9 Applications of Azure Quantum Elements
  • 5.3.10 Google DeepMind's GNoME and the Vertical Integration Play
  • 5.3.11 Meta's FAIR, OMat24 and the Open Catalyst Project
  • 5.3.12 Taking Materials Informatics In-House
  • 5.3.13 Offering In-Housed Operations as a Service
  • 5.3.14 Retrosynthesis Prediction
  • 5.3.15 Commercial Retrosynthesis Predictors
• 5.4 MI Consortia and Public-Private Initiatives
  • 5.4.1 NIMS and Materials Open Platforms (Japan)
  • 5.4.2 AIST Data-Driven Consortium (Japan)
  • 5.4.3 Toyota Research Institute and University Collaboration
  • 5.4.4 The Global Acceleration Network
  • 5.4.5 IBM Collaborations
  • 5.4.6 ChiMaD and the CMD Network
  • 5.4.7 The Open Catalyst Project: Crowdsourcing MI
  • 5.4.8 Materials Genome Initiative (MGI) — U.S.
  • 5.4.9 Materials Genome Engineering / National Materials Genome Project (China)
  • 5.4.10 Horizon Europe Materials Initiatives
  • 5.4.11 K-Moonshot
  • 5.4.12 Additional Initiatives and Research Centers
• 5.5 Corporate Initiatives in MI
• 5.6 Strategic Collaborations and Agreements 2024–2026
• 5.7 Geopolitics, Export Controls and MI
• 5.8 Applications of Materials Informatics
  • 5.8.1 Project Categories in MI
  • 5.8.2 Application Progression
  • 5.8.3 Materials Informatics Roadmap 2026–2036
• 5.9 Market Forecast and Outlook
  • 5.9.1 Market Forecast: External Materials Informatics Players (Provider Revenue)
  • 5.9.2 Market Forecast: Total MI Software & Services Market
  • 5.9.3 Forecast Data and Market Outlook
  • 5.9.4 Sensitivity Analysis: Bull, Base, and Bear Scenarios
• 5.10 MI Industry Player Data
  • 5.10.1 Lists of MI Players
  • 5.10.2 Full Player List — Commercial Companies (Confirmed Operational)
  • 5.10.3 Industry Leavers (Likely and Confirmed)

6 COMPANY PROFILES (53 company profiles)

7 RESEARCH METHODOLOGY

8 REFERENCES