Evolving Workflows

Current Demonstration Projects

The following projects are being developed as practical demonstrations of AI-assisted materials R&D workflows. The objective is to evaluate the usefulness of candidate-ranking methods, HTS planning workflows, literature-to-database pipelines, and AI-assisted experimental planning in real research scenarios.
* Note: These demonstrations are under active development and validation. Results shown are illustrative examples and should not be interpreted as experimentally validated outcomes.

Conceptual Materials AI R&D Lifecycle Flow

01
Define target properties
02
Build AI-ready dataset
03
Rank candidate compositions
04
Assess stability, cost & risk
05
Generate HTS matrix
06
Run focused validation
07
Active learning loop feedback
Electrochemical Demos

1. SOEC Anode / Oxygen Electrode Decision Map Demonstration

This demonstration explores how target conductivity, TEC, stability, cobalt reduction, and electrolyte compatibility constraints can be translated into candidate-ranking workflows and experimental validation plans.

Example Inputs:Target oxygen electrode conductivity (>200 S/cm), TEC limit (<12.5 x 10⁻⁶/K), cost & chemical stability constraints.
Example Outputs:Prioritized composition rankings, phase stability maps, and target validation campaign matrices.
Request Demo Details →
Supply Chain & Cost Demos

2. Critical-Mineral-Lean Perovskite Oxide Screening Demonstration

Explores active learning concepts to identify potential substitute compositions for functional ceramics, balancing raw material supply constraints and estimated material costs against baseline electrochemical indicators.

Example Inputs:Target electrochemical activity, element exclusion list, maximum raw material cost per kg.
Example Outputs:Alternative element substitution maps, cost-vs-performance Pareto fronts, and risk-rated composition candidates.
Request Demo Details →
Automation & HTS Demos

3. AI-Assisted HTS Campaign Layout Planner

Evaluates workflow methods to convert virtual candidate lists into structured experimental layout plans, planning synthesis matrices for robotic powder dispensers or liquid handlers.

Example Inputs:Selected candidate compositions, library layout constraints (e.g., 24-well or 96-well grid).
Example Outputs:HTS campaign layout plans (CSV/JSON templates), deposition boundaries, and planning configurations.
Request Demo Details →
Data Pipeline Demos

4. AI-Assisted, Human-Reviewed Literature-to-Database Pipeline

Demonstrates methods for converting published materials data into structured, traceable, AI-ready datasets. Evaluates extraction accuracy from PDF journal articles and patents into standard schemas.

Example Inputs:Unstructured research articles, PDF documents, or patent text files.
Example Outputs:Clean relational database tables (SQL/JSON), provenance-linked data records, and AI-ready training datasets.
Request Demo Details →
R&D Productivity Demos

5. Expert-Guided R&D Planning Assistant

This demonstration explores how AI can assist researchers in generating candidate lists, screening plans, measurement strategies, and next-step recommendations.

Example Inputs:Natural language R&D prompts, equipment specs, and safety/processing constraints.
Example Outputs:Experiment measurement checklists, custom laboratory data-entry schemas, and step-by-step processing recipes.
Request Demo Details →
Metallurgy & AM Demos

6. High-Entropy Alloy (HEA) Additive Manufacturing & Bypassing

Demonstrates physical-guided HEA discovery and patent-bypassing optimization for LPBF 3D printing. Evaluates thermodynamic solid solution stability, Pugh's ratio ductility, and volumetric energy density constraints.

Example Inputs:Target hardness (>450 HV), target toughness (>20 K1c), active patent (US-10829841-B2) limits, and LPBF print settings.
Example Outputs:Patent-free novel composition candidates (e.g., Cr lower limit bypass at 17.6%), Miedema mixing enthalpy and Omega calculation metrics, and optimal print window energy density validations.
Request Demo Details →
🛡️ Security and IP Protection CommitmentPlease note that the chemical formulas, performance ranges, and parameter constraints displayed above are purely for illustrative and demonstration purposes. In any actual consulting engagement or R&D project, the customer's core recipes, proprietary descriptors, and raw data are kept strictly confidential, isolated, and will never be shared or exposed to external entities.