Big Data, cloud & mobile computing, the Internet of things (IoT), and programmable microprocessors are some of the recent technology trends that have been altering the way engineers and scientists work, and have presented them with solutions as well as challenges. At the 5th MATLAB EXPO held recently, Jason Ghidella, Manager, Product Marketing for Simulink Platform and Controls, MathWorks delivered the keynote address titled 'Behind Today's Trends: The Technologies Driving Change', addressing some of these technologies. On the sidelines of the event, he spoke to us about the development of Simulink and MATLAB towards current technology trends, as well as the recent updates in the company's products towards the automotive sector.
SIMULINK & NEW TOOLS
Simulink is mainly used to understand and analyse complex systems by building and simulating blocks diagrams. It is used in all industries, with an exceptionally high degree of usage in the automotive sector, both by suppliers and OEMs. "A lot of Simulink's general capabilities for modelling, simulation and cogeneration are aimed at automotive customers," said Ghidella. He added that feedback is collected from key customers, and their requirements are incorporated into the tools, while releasing newer versions. One such example was a tool used for the calibration of engines, which was developed specifically for the automotive sector.
MathWorks invests heavily in R&D, and releases a new product every six months, once in March and the other in October. The most recent release is a testing product called Simulink Real-Time, which allows customers to build, run and test real-time applications, in the sense that a model can be taken from Simulink and incorporated into real-time prototyping hardware. Therefore, instead of putting the model into an embedded ECU, it can be applied onto a prototyping system and connected to the vehicle for testing. Simulink Real-Time can also be used to simulate hardware for testing, when it is not immediately available, and this is known as hardware-in-the-loop simulation. The final use of this tool is in the simulation of vision systems like cameras.
Another tool from the recent update is with reference to the management of data. While customers build models, especially in the automotive industry, a lot of design data associated with models is accumulated, said Ghidella. In order to efficiently manage this data, MathWorks introduced Simulink Data Dictionary to allow design data to be stored in a separate repository which can be configured. The data can be referred back to any time and be used for further development activities.
Over the last few years, MathWorks has developed ways to take MATLAB software mobile. It includes MATLAB Mobile, MATLAB Desktop and MATLAB on cloud. MATLAB Mobile can be used on Apple and Android devices by connecting to a work computer or the MathWorks cloud network and accessing the necessary files. The second method is to connect to MATLAB on cloud or the Amazon EC2 cloud, which registered users have access to. Users can also use a public desktop that does not have the MATLAB software installed by utilising the MATLAB distributed computing server or through the cloud server.
CURRENT TECHNOLOGIES AND THE ROLE OF MATHWORKS
Ghidella observed that trends are always changing, but there is something beneath them that drives the trends, and that factor is constant. There are technology transformations that are driving trends, and these include algorithms, hardware, connected systems and the fact that people want to compute. These four technology transformations are especially driving trends in areas of Big Data and IoT, which create a completely connected environment.
Global connectivity is what technology is trying to achieve currently, and this can be segregated into three parts, explained Ghidella. First is the connecting of places by the various methods of transportation, and then the connectivity of people through mobile devices and computers. The final part of global connectivity is the connectivity of all things, meaning all devices and systems need to be synched with one another and need to communicate to each other.
Once devices connect with each other, they create large amounts of data that can be put to endless use, when collected and analysed. In terms of the automotive industry, such data can help in developing and updating technologies related to connected and autonomous vehicles. Data transfer between vehicles and infrastructure is also something that is becoming more and more a reality and this will also benefit from collection and analysis of appropriate data.
The role of MathWorks in all these trending technologies is that through specific tools, customers will be able to carry on data analytics and process the large set of data. The tools can be used for things like parallel computing, which can be used to process the data. MathWorks has also developed specific algorithms for machine learning in order to stream large amount of data and put it to use in various devices, Ghidella explained.
FUTURE OF MODELLING
There seems to be a gap between models/ prototypes and the final product, and this is something that every company offering simulation and modelling solutions is looking into. Ghidella said detailed models need to be evolved in the early stages of the development process in order to carry out simulation and make corrections. MathWorks also offers additional physical modelling capabilities in its products so as to enable modelling in the earlier stages. A main feature that MathWork's tools provide is information and data from previous models, which help in developing and improving future variants and models of a product.
Ghidella signed off by saying that physical modelling and prototyping cannot be done away with completely, but can only be reduced significantly in the future.
Text: Naveen Arul