Technical computing software vendor MathWorks is broadening the scope of its MatLab and Simulink by bringing in direct support for low-cost development hardware such as the Arduino, Lego Mindstorms, and Beagle Board products, and positioning the company potentially to take advantage of interest in the Raspberry Pi.
MathWorks senior engineer Mark Walker says the company has bundled a number of the tools needed to generate code to run on boards such as the Arduino family into the core product. “Historically, people have always been able to generate code for hardware but it required a large chain of tools," he says. “For users in education, it can now be done in one tool.”
The generated code runs natively on the target board but, Walker says, “you can interact with the Simulink model in the same way as you can with a software simulation”.
Although the current offering does not target the Raspberry Pi, targets such as the Beagle Board use a similar ARM processor. Walker said the company is likely to add support: “There is a lot of UK demand for the Raspberry Pi and we think that will build internationally as well.”
To provide more Tanya Morton, applications engineering manager, said the company noticed that a number of the features in the core tool and in its add-on toolbox often went unnoticed by users. The decision to promote apps using the company’s website is intended to show-off ready-made examples and functions in MatLab, and to let users share those that they have made themselves.
“We find that people have been using the toolboxes for a long time but haven’t found some of the sophisticated user interfaces that are already in there,” says Morton.
Other changes for the mid-year 2012b release of MatLab and Simulink include a number of user-interface enhancements that are meant to simplify the creation and management of large-scale simulations.
Walker adds: “The scale of systems that people model has gone up pretty rapidly. And the models are used between companies to communicate specifications.”
To make the models cleaner, the latest release uses automates the placement of routes between signals to avoid the need for users to clean up diagrams and to make it easier to navigate through the model hierarchy. “You need hierarchy as it’s pretty hard to work with a very large model in one go,” Walker avers.