Québec City, Canada | September 29, 2026 – TACTUN participated in the International Materials Applications & Technologies (IMAT) Conference, organized by ASM International, where leading researchers and engineers discussed innovations shaping the future of material science and engineering.

TACTUN CEO and Co-Founder Rafayel Ghasabyan presented “‘Vibe-code’ Your Materials Testing Machine: AI-Native Configuration and Automation with MCP” as part of IMAT’s Artificial Intelligence and Materials Informatics technical session.

From Programming Machines to Expressing Engineering Intent

The presentation explored a new interaction model for materials testing systems: instead of manually programming every test sequence, machine configuration, and reporting workflow, engineers can increasingly describe what they want the machine to accomplish, while AI translates that engineering intent into a structured configuration.

A central element of the approach is the Model Context Protocol (MCP), which can provide AI systems with structured access to machine capabilities, configuration tools, and data.

This creates the possibility of an engineering workflow in which natural-language instructions can be transformed into test sequences, parameters, operational logic, data processing, and reporting workflows.

AI Above the Real-Time Control Layer

A key technical principle presented at IMAT was the separation between AI-assisted engineering and deterministic machine control.

AI can assist engineers in creating and modifying machine configurations, but safety-critical functions and closed-loop motion and process control remain within the deterministic real-time control layer.

This distinction is particularly important for materials testing, where repeatability, precise timing, traceability, and machine safety cannot depend on the probabilistic output of a language model.

The presentation discussed practical applications including generating fatigue profiles, configuring ramp-and-dwell sequences, adapting test parameters, and creating standards-oriented test and reporting workflows.

“The goal is not to let AI replace deterministic control. The opportunity is to let engineers express what they need in engineering terms and have AI translate that intent into a structured machine configuration, while the real-time controller remains responsible for precise and safe execution.”

Toward AI-Native Materials Testing Machines

The broader idea presented at IMAT represents a transition from code-centric machine development toward intent-driven engineering.

As machines expose their capabilities through structured interfaces such as MCP, domain experts may increasingly be able to configure sophisticated testing systems without manually developing every element of the underlying application.

For materials testing, this could reduce the engineering effort required to introduce new test methods, modify existing procedures, automate reporting, and adapt equipment to changing requirements.

At the same time, the architecture preserves a clear boundary between the probabilistic AI layer and the deterministic real-time control layer — an important requirement for bringing generative AI into physical machinery.

The presentation was part of IMAT 2026’s broader technical discussion around artificial intelligence, materials informatics, automation, and emerging approaches to materials research and engineering.