Design tools tackle big-data challenge

  • 30-Oct-2012 04:57 EDT
aetrdspace.jpg

Mazda used MathWorks tools to design and calibrate its Skyactiv engine.


Design and validation processes that include modeling, simulation, and hardware-in-the-loop testing are creating larger and larger volumes of data, making it more challenging for engineers to analyze and understand the nuances of their designs. Tool providers are ramping up their efforts to help these designers manage data.

Understanding the growing amount of data was an issue that popped up in many areas during the recent 2012 SAE Convergence conference. Even a keynote speech on IBM’s Watson computer that gained fame by beating humans in “Jeopardy” included a comment about the vast amount of data that’s being generated today.

“Ninety percent of the data available now has been created in the last two years,” said Rob High, CTO of IBM’s Watson computer program.

Tool providers on the show floor demonstrated technologies designed to help engineers get the most from the data they create. For example, dSpace showed off its Synect data management tools for model-based design.

“Complexity is rising, so there’s a greater need to manage data at the back end,” said Mahendra Muli, New Business Development Director at dSpace. “There’s a big tool chain around model-based design, so there are a lot of data types. We’re making it easier for users to find and manage data.”

At Convergence, Mazda divulged that it used The MathWorks tools to develop and calibrate its Skyactiv engine technology. The company said it couldn’t achieve the performance and fuel economy ratings without using model-based design.

“Finding an optimal calibration setting in a search space of five or more dimensions is difficult even for experienced engineers, so we could never be certain that we had found the best possible settings,” said Shingo Harada, Assistant Manager at Mazda.

“Models let Mazda see more of what was happening inside the engine,” added Wensi Jin, Transportation Industry Manager at The MathWorks. “Once they collected the data, Mazda needed to calibrate the system with a model-based calibration toolbox.”

Other companies are also beefing up their efforts to help engineers manage the huge volume of data that is created when complex systems are run through many iterations of tests and simulations.

“Our latest version adds a lot to help engineers search through data,” said Marcella Haghooie, Senior Field Applications Engineer at Applied Dynamics International. “There are new ways to compare results and compare results from different runs.”

Some companies are partnering to ensure that users will have compatible tools that complement each other and help engineers do more tasks in compatible environments. That can eliminate some of the challenges that come when tools create data in different formats.

“We are working with IBM, including our test products with their Rational software development platform, so users can have end-to-end quality management,” said Noah Reding, Automotive Product Manager at National Instruments.

Share
HTML for Linking to Page
Page URL
Grade
Rate It
3.83 Avg. Rating

Read More Articles On

2015-06-09
To reduce cost and improve performance in engines, Honeywell is offering a software tool to embed virtual sensors in engines, using first-principles models that resides in the engine control unit.
2015-06-23
Elektrobit releases an easier-to-use toolkit that enables development of multimodal user interfaces with advanced graphics, plus voice-, touch-, and gesture-based commands, for sophisticated human-machine interface (HMI) systems.
2015-06-09
The AUTOSAR standard is all about improving embedded electronic systems in vehicles, but did you know that it also promotes the early testing of software code? The early testing and validation of embedded software reduces the number of ECU prototypes and additional test systems needed down the road, and it benefits the whole development process.
2015-06-09
New surface scanning technology is being utilized by rFpro to produce digital road models with unprecedented accuracy and speed to help make autonomous driving safer.

Related Items

Article
2015-05-26
Technical Paper / Journal Article
2003-10-27
Article
2015-06-05