Generative AI: The future of product design

Generative AI: The future of product design

By Raghavendra K.A., SVP and Global Head of Engineering – Integrated Product Development, Engineering Systems & IoT, Infosys

Add product design and development to gen AI’s list of unending possibilities. While the sky may well be the limit, a leading consulting firm estimates that gen AI could add $60 billion in value by way of productivity in product research and design.

From design to delivery

The role of generative AI is not limited to design and development, but encompasses the entire product lifecycle, from market research to concept creation to testing and improvement. AI has been playing an important role in product – across these areas – in three broad phases. In the first phase, Knowledge Based Engineering (KBE), a branch of deterministic AI, played an important role in improving productivity and quality. In the second phase, the use of machine learning and deep learning concepts have helped us to conceive smart and connected products. In the third phase, which is the current phase, the age of generative (gen) AI and Large Language Models (LLMs), we are able to gather and analyse market data from a huge number of sources, and uncover insights, such as unmet needs and emerging opportunities, in a fraction of the time and cost of traditional research.

At the idea and design generation stage, beyond meeting functional specifications like performance, sustainability, and compliance, gen AI tools can help organisations connect more closely with consumers by envisioning boundary-breaking products that are out of reach for conventional development. A group of individual product designers can also share their designs among themselves on a gen AI-powered collaboration platform to understand the full range of possibilities, accommodate diverse perspectives, and collectively arrive at a superior base design. Next, they can iterate the design(s) to quickly create variants and improvements, shortening time-to-prototype. AI-powered real-time feedback loops provide continuous insights to help product teams quickly identify strengths and improvement areas (think risks, design flaws, safety issues and other errors) and finalise prototypes faster. Armed with a gen AI-visualised model of a new product, designers can approach potential users for feedback and then take the proposition forward with internal stakeholders. They can continue to use the technology to make further improvements to the design and subsequently add the detailing and engineering concepts to make it ready for manufacturing.

Given how quickly gen AI is evolving, experts believe that it will extend its role to post-design activities, such as engineering and manufacturing, sooner rather than later. For example, gen AI tools will streamline and speed up handover from design to manufacturing to operations & maintenance – they are already converting product ideas into CAD models – and assessing whether a particular design can be manufactured with a factory’s existing production assets. From selecting (or even designing) the right materials to optimising parts and suggesting ways to make a design more manufacturable, generative AI is likely to do it all. It can also envision potential operational, maintenance and warranty issues.

Designed by AI, decided by humans

Doubtless, gen AI can transform product design and development to deliver myriad benefits: productivity, cost efficiency, faster innovation, and highly optimised designs, to name a few. But it does not eliminate the human factor. Ultimately, human product designers and managers will decide which designs to use, and where and how to use them, based on various considerations such as strategic alignment and business value. There needs to be a human in the loop to oversee the working of the technology, especially its adherence to responsible AI principles: making sure the training data is fair, complete, and accurate to generate positive design outcomes; ensuring that the designs don’t cause safety issues, or violate intellectual property rights, etc. Last but not the least, choosing designs that are pleasing to our aesthetic and design sensibilities is something that only we humans can do for now.

AIGenAIInfosys
Comments (0)
Add Comment