HANNOVER MESSE 2019, 01 - 05 April
switch to:
Predictive Maintenance

MathWorks R2018a gives designers more options

The latest release comprises new data analytics options for predictive maintenance. Also included in the package: deep learning and autonomous driving.

26 Mar. 2018
MathWorks R2018a gives designers more options (photo: The MathWorks, Inc.)

It is not only engineers who usually discover MathWorks prior to university: as state-of-the-art software for numerical calculations and modeling dynamic systems. Release 2018a has now once again significantly expanded the range of functions provided by MATLAB and Simulink to include, among other things, a predictive maintenance toolbox and new deep learning options.

Predictive Maintenance Toolbox can handle local data, as well as data from external sources, such as the cloud and distributed systems. The reference examples for engines, transmissions, batteries, and other machines, delivered as standard with the software, help users develop their own predictive maintenance and condition monitoring algorithms. At the same time, Neural Network Toolbox has a new add-on that accelerates the relevant machine learning processes, implements optimization techniques, and supports, for instance, networks with DAG topologies (directed acyclic graphs) and pre-trained networks such as GoogLeNet, ResNet, and SegNet. In addition, the software can now also generate C code for deep learning networks on Intel and ARM processors and output CUDA code. One thing is therefore for sure: If predictive maintenance projects fail , it is not down to the developers or the design.

What is more, the Vehicle Dynamics Blockset ensures the availability of a complete, customizable standard 3D model architecture for automotive simulations, from drives, steering, suspension, brakes, and other vehicle components through trees, streets, and traffic signs.