The rise of deep learning technology has been a key catalyst for the introduction of artificial intelligence (AI) to automotive engineering. From advanced topology optimization used in mechanical design to state-of-the-art perception algorithms used in active safety systems, many parts of vehicles today have been touched by the use of Artificial intelligence. After all, most perception systems today have the basis of their detection algorithms in some form of supervised learning.
Already, we are witnessing generative AI being used in the process of jumpstarting structural and electrical designs, providing engineers with valuable initial design possibilities when seeded with initial design constraints. It is no surprise, therefore, that generative design can result in a much more creative and more ‘out-of-the-box’ design. Already today, CAD/CAE vendors such as Ansys and Autodesk have started offering tools that can use this new development.
However, in design, automotive designers typically need to balance the trade-offs between performance, cost, and manufacturability when designing a vehicle, in addition to safety and durability. The rigor from those criteria typically feeds useful real-world data that will provide useful context for the designed parts and systems. From generative AI comes generative design, a new framework that promises to revolutionize the engineering design process.
Autonomous vehicles are gradually expected to come to reality within the next decade or so.
The availability of additive manufacturing technology allows for complex electrical and mechanical designs engineered and created by generative AI. As a technology, additive manufacturing is rapidly reimagining the rules of how metal parts can be made, thanks to a new design language - Design for Additive Manufacturing (DfAM).
This language was developed to take advantage of the enormous design freedom that comes with 3D printing, and true to form, DfAM allows manufacturers to unlock the value of this new approach - from highly complex geometry to the possibility of assembly consolidation by the use of generative design tools to create parts that would otherwise be too difficult or costly to justify with traditional manufacturing methods. In addition, this new way of thinking spurs the creation of innovative products and reduces waste during fabrication.
Meanwhile, from the electrical vehicle architecture perspective, the industry stands on the verge of convergence with edge computing, machine vision, and 5G-connected vehicles, culminating in AI-defined vehicles.
For the past few years, we have witnessed the growing intelligence of vehicles and mobile edge computing with recent advancements in embedded systems, navigation, sensors, human-machine interfaces, and data analytics. What started with Advanced Driver Assistance Systems (ADAS), such as emergency braking, backup cameras, adaptive cruise control, and self-parking systems, is starting to form the basis of the next generation of highly automated driving features.
Autonomous vehicles are gradually expected to become a reality within the next decade. Following the introduction of the six levels of autonomy defined by the Society of Automotive Engineers (SAE). These levels range from no automation and conditional automation (human in the loop) to fully automated vehicles. With each increasing level of automation, the vehicle will be more capable of taking over more functions from the driver for the various Operational Design Domains. ADAS mainly belongs to Level 1 and Level 2 of automation, depending on whether lateral control is involved.
Truly, the convergence of 5G connectivity, edge computing, advances in computer vision, and deep learning will be key to delivering the AI-defined vehicle. The continued development of standard and interoperable technologies such as V2X, more powerful edge computing, and onboard inferencing accelerator technologies will deliver on the promise of low latency, energy-efficient, low-cost, and safety benefits for this exciting next generation of vehicles. Artificial intelligence will certainly change how we design and build vehicles, as well as how they will work for us in the future.
