Product design needs AI transformation


This is a daunting task, but Zapf said that artificial intelligence (AI) technology can provide support by capturing correct data and guiding engineers in product design and development.

No wonder a survey conducted by McKinsey & Company in November 2020 showed that more than half of organizations have adopted AI for at least one function, and 22% of respondents indicated that at least 5% of their company-wide revenue is attributable to AI. In the manufacturing industry, 71% of respondents who adopted AI saw their income increase by 5% or more.

But this is not always the case. Zapf said that AI, which was once “rarely used in product development,” has undergone a development in the past few years. Today, technology giants such as Google, IBM, and Amazon, known for their AI innovation, “have set new standards for the use of AI in engineering and other processes.”

Katrien Wyckaert, Industry Solutions Director of Siemens Industrial Software, said: “Artificial intelligence is a promising field of exploration that can significantly improve the user experience of design engineers and collect relevant data during the development of specific applications.”

As a result, people have a higher and higher evaluation of a technology that is expected to simplify complex systems, bring products to market faster, and promote product innovation.

Simplify complex systems

Renault is an excellent example of AI’s comprehensive innovation in product development capabilities. In response to growing consumer demand, the French automaker is equipping more and more new models with automatic manual transmission (AMT), which is similar in performance to an automatic transmission but allows the driver to use button commands Perform electronic shifting.

AMT is very popular among consumers, but designing them can present huge challenges. This is because the performance of AMT depends on the operation of three different subsystems: electromechanical actuators for shifting gears, electronic sensors that monitor the state of the vehicle, and software embedded in the gearbox control unit that controls the engine. Due to this complexity, it may take up to a year of trial and error to define the functional requirements of the system, design the actuator mechanism, develop the necessary software and verify the entire system.

In order to simplify the AMT development process, Renault chose Simcenter Amesim from Siemens Digital Industries Software. Simulation technology relies on artificial neural networks, which are AI “learning” systems loosely modeled on the human brain. Engineers only need to drag and drop, connect the icon to create a model graphically. When displayed on the screen as a sketch, the model illustrates the relationship between all the various elements of the AMT system. In turn, engineers can predict the behavior and performance of AMT and make any necessary improvements early in the development cycle to avoid problems and delays later. In fact, by using virtual engines and transmissions as an alternative when developing hardware, Renault has successfully cut its AMT development time by nearly half.

Speed ​​without sacrificing quality

Similarly, emerging environmental standards have also prompted Renault to rely more on artificial intelligence. In order to meet the emerging carbon dioxide emission standards, Renault has been committed to the design and development of hybrid vehicles. However, the development of hybrid engines is much more complicated than found in vehicles with a single energy source, such as conventional cars. This is because hybrid engines require engineers to perform complex tasks, such as balancing the power required by multiple energy sources, choosing from multiple architectures, and checking the impact of the gearbox and cooling system on the vehicle’s energy performance.

Renault’s head of simulation, Vincent Talon, said: “To meet the new environmental standards for hybrid engines, we must thoroughly rethink the architecture of gasoline engines.” He added that the problem is that careful inspection “may affect fuel consumption and pollutant emissions. “Dozens of different actuators” is a long and complicated process, and it is even more difficult due to tight time.

“Today, we obviously don’t have time to arduously evaluate various hybrid powertrain architectures,” Talon said. “Instead, we need to use an advanced method to manage this new complexity.”

For more information about AI in industrial applications, please visit

download Full report.

This content was produced by Insights, the custom content department of MIT Technology Review. It was not written by the editors of MIT Technology Review.


Source link