Join top executives in San Francisco July 11-12 to hear how leaders are integrating and optimizing AI investments for success. Learn more
Toyota Research Institute (TRI) presented an innovative Generative Artificial Intelligence (AI) tool that aims to improve the creative process of vehicle designers. The tool allows designers to generate design sketches via text prompts, incorporating precise stylistic attributes such as “stylish”, “SUV-like” and “modern”. Additionally, designers can optimize quantitative performance metrics to create an initial prototype sketch.
The company said this innovation will allow designers to explore their creativity while ensuring efficient and effective design development.
TRI researchers have also published two cards describing how the developed technique can be incorporated into other text-in-image AI models. These documents shed light on the instrument’s imaging process.
The team combined the principles of optimization theory, widely used in computer-aided engineering, with text-in-image generative AI. As a result, the algorithm allows designers to optimize engineering constraints while preserving their text-based stylistic suggestions for the AI process.
Improving image generation
Designers can now implicitly integrate vehicle constraints such as aerodynamic drag, which directly affects fuel efficiency, and chassis dimensions such as ground clearance and cabin size, which affect handling, and security to improve image generation.
“Current text-to-image AI tools are primarily focused on adhering to the designer’s text-based stylistic guidelines when generating potential images,” Avinash Balachandran, director of Human Interactive Driving (HID) told VentureBeat. ) at the Toyota Research Institute. “Our technique allows users to explicitly incorporate and optimize over-engineering constraints such as drag or ride height while generating images that adhere to the designer’s stylistic guidelines.”
Balachandran said such techniques could speed up the creation of new designs by balancing the trade-offs between aesthetics and engineering more quickly and efficiently.
“Any designer can use AI tools for inspiration, but these tools cannot handle the complex engineering and safety considerations that go into actual automotive design,” he added. “To build safe and reliable vehicles, our designs must meet engineering requirements. Adding constraints to Generative AI essentially allows the user to add guide rails to Generative AI projects.
Optimizing vehicle design through generative AI
Balachandran told VentureBeat that the project began about a year and a half ago, driven by advances in text-to-image AI tools that allowed users to enter a prompt and in response generate an image that aligns with style guide provided.
“Our vehicle designers at Toyota have told us how one of the challenging parts of the design process was for them to find inspiration for new designs,” he explained. “They also told us that the process of iterating back and forth between design and engineering to produce a design that is not only aesthetically pleasing but also has the desired engineering performance and safety measures was difficult.”
According to Balachandran, designers and engineers typically come from different backgrounds and have different ways of thinking. As a result, when a designer creates a design, he usually fails to meet the initial engineering requirements, resulting in substantial collaboration with the engineering team to arrive at an optimal solution.
This iterative process, coupled with the inherent tension between design and engineering, contributes to the extended life of the design.
“The inspiration for this technique and these tools was not only to stimulate creativity but also to shorten the iteration cycle between engineering and design,” said Balachandran.
Incorporate different data streams
Toyota said that during ideation sessions with the designers, one idea that resonated with them was the concept of an “AI assistant” that proposes new designs by leveraging multiple different data streams. This gave birth to the idea of integrating generative AI into a tool that incorporates different data streams, including engineering constraints, to generate innovative designs.
“By integrating Generative AI technology, we found that designers were able to focus on identifying important design constraints and design aspects with confidence that the engineering constraints were met,” Charlene Wu, senior director of Human- Centered AI (HCAI) division of the Toyota Research Institute, he told VentureBeat. “We believe our tool will allow them to spend more time on the part of the design process they value the most and where they can add the most value.”
What is the future of Toyota?
The company announced that while the technology is currently in the research stage, they are collaborating with teams within Toyota to integrate this tool into the vehicle design and development process. TRI has stated that it will continue research to improve the quality of life of individuals and society.
“The hope is that, using this tool, vehicle designers around the world can expand the power of design ideas while dramatically improving the speed of design development,” said Balachandran. “Generative AI is a powerful new tool, and in our many research areas, we are exploring how to responsibly harness it so it can amplify people.”
VentureBeat’s mission it is to be a digital city square for technical decision makers to gain insights into transformative business technology and transactions. Discover our Briefings.