April 30, 2026
News
MathWorks has unveiled Release 2026a (R2026a) of its MATLAB and Simulink product families, incorporating advanced AI functionalities designed specifically for embedded systems development. This latest version features the Simulink Copilot to bolster Model-Based Design and Polyspace Copilot for enhanced embedded software code analysis, empowering engineering teams to sustain productivity, rigor, traceability, and repeatability throughout their designs.
MathWorks is advancing AI in engineering in two key ways:
Firstly, the company is embedding copilots within the existing environments that engineering teams are familiar with, such as MATLAB Copilot, Simulink Copilot, and Polyspace Copilot.
Secondly, MathWorks integrates MATLAB and Simulink functionalities into agentic workflows via the MATLAB MCP Core Server and MATLAB Agentic Toolkit. This enables teams to comprehend designs more swiftly, resolve software issues earlier, and implement development and verification workflows with greater consistency.
Simulink Copilot, founded on user-defined models, processes, and comprehensive MathWorks documentation, offers guidance specifically tailored to the engineering environments in use. It can generate model explanations, answer queries regarding model functionality, and assist users in locating applicable blocks and subsystems. By pinpointing issues, recommending solutions, and directing subsequent steps, the copilot facilitates faster progression in design tasks. Additionally, teams can execute standardized tasks that promote uniform development and verification methods.
R2026a introduces the Polyspace Copilot and Polyspace as You Code. The Polyspace Copilot assists engineers in deciphering static analysis results, while Polyspace as You Code enables developers to verify C and C++ coding standards and identify potential coding defects and vulnerabilities in real-time, including code generated with AI-supported tools.
In addition, R2026a brings three enhancements across the Polyspace suite: a new desktop application for centralized configuration and results management, improvements to Polyspace Bug Finder with customizable checkers and coding standards, and enhanced capabilities for runtime error analysis within Polyspace Test.
The key updates in R2026a across the MATLAB and Simulink families include:
- MATLAB Course Designer: A new tool enabling educators to create courses, courseware, labs, and assessments with MATLAB and Simulink.
- Simulink FMU Builder: This new product facilitates the creation of standalone Functional Mockup Units from Simulink models and C or C++ code for improved model exchange and integration.
- MATLAB: Engineers can now generate and share interactive webpages featuring visualizations without needing to install MATLAB, and manage Python environments more effectively.
- Simulink: Streamlined access to commonly used actions via simplified, task-oriented context menus. Users can now simulate C and C++ code within models, without language constraints or the need for additional wrappers.
- Wireless Network Toolbox: This feature allows users to design, simulate, analyze, and visualize wireless communication networks for comprehensive evaluations of overall system performance.
- MATLAB Test: Users can generate initial tests, equivalence tests, and tests from command history through MATLAB Copilot, including capabilities for executing context-relevant tests to minimize unnecessary runs.
- Mapping Toolbox: Facilitates in-depth geospatial analysis with 3D building visualizations and image overlay functions.
- Signal Processing Toolbox: Offers new tools for the design and analysis of digital filters, enhanced label capabilities for time-frequency data, and tools for extracting signal features.
“Engineering teams now have access to capabilities enhanced by generative AI, and it is crucial for leaders to ensure these translate into meaningful engineering and business outcomes,” states Avinash Nehemiah, Head of Product Management and Marketing, Design Automation at MathWorks. “Improvements in productivity must not compromise rigor, traceability, or trust. MathWorks is dedicated to providing grounded AI tools for engineering that allow teams to innovate swiftly while maintaining the discipline and confidence essential for developing complex engineered systems.”
For further details, visit R2026a Release Highlights.
Tiera Oliver serves as the assistant managing editor at Embedded Computing Design, where she specializes in web content editing, product news, and story creation. Additionally, she oversees content development for ECD podcasts, including Embedded Insiders.
Leveraging her expertise in journalism and content management, Tiera coordinates with editors to help ensure high-quality output across web, print, and multimedia formats. She manages a variety of projects, aids in the production of digital magazines, and conducts in-depth conversations with industry leaders for engaging and insightful discussions on company podcasts.
A graduate of Northern Arizona University, Tiera earned a degree in journalism and political science. She also gained experience as a news reporter for the student-run publication, The Lumberjack.
