In an increasingly complex world, what with the growing demand for superior technologies in all spheres of life, scientists, researchers and engineers working on advanced and complex technologies are constantly looking for faster and smarter ways to break down and manage the resulting complexity. At the recently-held MATLAB EXPO, the fourth edition in India, Jim Tung, MathWorks Fellow, delivered the keynote address titled ‘Embracing Complexity’. While MathWorks products are essentially used to help avoid complexities, what justifies Tung’s subject? We find out.
Some of the possible grand challenges the automotive industry faces today include zero automotive traffic fatalities, minimised injuries, and significantly reduced traffic congestion and delays. To address these challenges, it is important to understand what needs to be built. But solutions to these critical challenges are often complex, said Tung. Today, it is about managing complexities, and at MathWorks, the process is usually four-fold: collaborate, model, simulate and automate.
“We are at a stage today in automotive engineering, where you’ve got different parts of the vehicle such as the engine or transmission that are fairly monolithic. Then there are other things that are more distributed across the vehicle like chassis stability or body electronics. Certain things are complex because of their functionality, like engine management, and some things are complex because of the distribution,” Tung explained.
Take for example, Adaptive Cruise Control. Although it seems like one function, in reality it is a complex communication and collaboration system of the Engine Management Control, Electronic Stability Programme and the Braking System. Similarly, Crash Avoidance, Adaptive Front Lighting Systems, Passenger Detection or Driver Drowsiness are aspects of complexity that need to be managed.
Developments like AUTOSAR is a result of these complexities, where Tier I suppliers and OEMs are working towards standardisation of automotive software architecture, Tung noted.
Many answers to the complexities, like mentioned above, could be found through proper data analysis. With features like collision avoidance and lane departure warning systems being increasingly featured in premium vehicles, the type of sensors and data are getting more diverse. From the MathWorks standpoint, today’s world of data analysis means being able to support all the steps – observation, organisation, understanding and application – in a fluid and flexible manner. And that is where MathWorks historically has been very well known for.
“One aspect that is increasingly getting important is the ability to model the various physical aspects of the system of the car, be it mechanical, electromechanical or hydro-electromechanical. It’s important to be able to do all of these things at one place. To let the engineers figure out where is the most flexible place to do a design change is increasingly important,” said Tung.
Tung was essentially pointing at the necessity to change the way data analysis is being done. “What we are seeing across industries is that the traditional way of data analysis has to change. Increasingly, the world of data science or data analytics is changing.”
Autonomous vehicles offer tremendous new opportunities for MathWorks. Although in its infancy, automated driving is increasingly being looked upon by the industry, primarily in mature western markets. Today, automated driving is a long-term strategy for most, with fully automated vehicles likely to become a reality only by 2025, or thereabouts. Any level of automated driving would require adequate and correct communication mediums to be developed. Concurrently, it would also be important to ensure the right amount of data analysis is carried out. Autonomous vehicles, or automated driving, would also be dependent on the proper development of vehicle to vehicle (V2V) communication and vehicle to infrastructure (V2I) communication.
Another thing that is interesting in the automotive space is the changing relationship between the OEMs and the suppliers. “The classical approach in most traditional automotive domains like the powertrain or chassis is that the OEM sets the requirements of what the need is, and the Tier I develops and gives back a fairly complete sub-system back to the OEM. But when we look at the newer propulsions technologies – hybrids, fuel-cells, electric vehicles – you see the OEMs not wanting to offload these technologies to a Tier I.”
With the marketplace getting increasingly complex, and with newer challenges facing the industry, Tung believes MathWorks would be able to deliver products that would help accelerate the pace of innovation and development.
Text: Deepangshu Dev Sarmah