A new engineering software release adds AI co-pilots, faster verification tools, and workflow automation to help embedded teams design, test, and deploy systems with greater efficiency.
MathWorks has launched Release 2026a of its MATLAB and Simulink product families, introducing new AI-powered copilots to accelerate embedded systems development while preserving engineering rigour, traceability and repeatability. The update adds Simulink Copilot and Polyspace Copilot, alongside broader workflow enhancements spanning design, verification and production.
The latest release focuses on embedding trusted AI directly into engineering environments already used by development teams. Simulink Copilot is designed to support Model-Based Design workflows by helping users understand model behaviour, generate explanations and quickly locate relevant blocks or subsystems. It can also assist in identifying issues, suggesting fixes and guiding next steps during development.
The key features are:
- Simulink Copilot for model guidance
- Polyspace Copilot for code analysis
- Real-time C/C++ rule checking
- FMU export for integration workflows
- AI-assisted automated test generation
For software verification, Polyspace Copilot analyses static code-checking results to help developers interpret findings and resolve defects faster. A new “as You Code” capability checks C and C++ coding rules, vulnerabilities and defects as code is written, including software generated through AI-assisted coding tools. This early-stage feedback is intended to improve software quality and reduce downstream debugging effort.
MathWorks said the release is built around grounded AI, meaning responses are based on user models, defined workflows and official documentation rather than generic outputs. This approach is intended to give engineering managers greater confidence that productivity gains do not compromise compliance or design integrity.
Beyond AI features, the update adds several new tools. A course design platform helps educators create training content using MATLAB and Simulink. A Functional Mockup Unit builder enables standalone FMU creation from Simulink models and C/C++ code for system integration tasks. Engineers can also build interactive webpages with visualisations without requiring MATLAB installation.
Additional improvements include streamlined Simulink context menus, integrated simulation of C/C++ code within models, wireless network simulation tools, expanded geospatial analysis functions, and upgraded digital filter design apps. MATLAB Test now uses AI assistance to generate starter and equivalence tests while reducing unnecessary execution.
The 2026a release signals a stronger push toward AI-assisted engineering software, particularly for embedded development teams seeking faster iteration cycles without sacrificing validation discipline.
Click here for the original announcement.