The nature of product design is changing rapidly. Customers expect sophisticated customization, while disrupted supply chains require alternate sources for materials and components. Manufacturers are under increasing pressure to maximize agility and minimize waste. And products themselves are no longer static objects: increasingly, the “end product” is in fact something that can continue to evolve and grow in functionality after it’s in the customer’s hands.
All these changes create new challenges for designers, engineers, and manufacturers. But technology offers both solutions and added opportunities in the space. AI-enabled digital tools will support a new generation of customer-centered, resilient, and sustainable products.
Design was traditionally driven by human experience and creativity, with technology playing a limited role via discrete tools. But recent tech advancements have started to dramatically shift the role of technology in the design process. Look at consumer insights. Customer feedback is one of the most significant inputs for a design process, but it can be difficult to get insights that are current and relevant for a specific product. Now, advanced natural language processing (NLP) and knowledge graph-driven techniques let designers capture relevant customer insights with higher accuracy and much, much faster. In some cases these technologies enable near real-time feedback from existing products that can feed into future product updates or evolutions.
Augmented and virtual reality also have a role to play, letting designers visualize physical products at a much earlier stage of the design process. Simulation and analysis platforms allow for evaluation of potential designs much earlier as well. Meanwhile, deep neural networks show great promise in augmenting designers by generating new design ideas. We’re working with all of these technologies, exploring what’s possible for the future of product design.
In one of our efforts, we’ve developed a system that hosts a suite of AI design assistants to help fashion designers in their journey. To start, the Creative Design Assistants Platform for Fashion (CDAP-F) includes a Consumer Insights Assistant. This assistant incorporates information from sales and trend analysis of apparel. For example, sales data may suggest that a specific color leads to more purchases. Designers can use such insights as they consider new designs.
Other assistants within the platform help designers easily explore a wider range of design variants. One, the Apparel-Style-Merge assistant, lets designers combine elements from multiple pieces of apparel to create new designs. Another, the Apparel-Style-Transfer assistant, makes it easier to explore different styles, colors and patterns. All of these assistants use variants of deep neural networks to help augment the capabilities of human designers.
An early user evaluation shows that the platform's good quality, unique design options significantly enhance creativity. Senior faculty members from the Department of Fashion and Textiles at an Indian university also confirmed these results. What’s more, the platform reduces design generation time from 15 mins to 2-3 minutes. This offers great potential for apparel design firms – they have to cater to the fast-changing fashion industry, where the costs of falling behind trends and consumer preferences are significant. An estimate done by ShareCloth finds that about 30% of apparel was never sold in 2018. Solutions like CDAP-F cut down design time while making it easier to keep up with customer insights and preferences, giving the generated dresses a better chance of success - contributing to sustainability by reducing unsold inventory. You can read more about AI-Assisted Apparel Design at DeepAI.
This isn’t the only way we can apply AI solutions to design challenges. We can also use AI-driven generative design to create highly customized products or parts for individual customers, while guaranteeing compliance with key design requirements.
Think about the auto industry. Buyers increasingly want vehicles that are customized to their preferences, but automakers need to ensure that any custom options meet safety standards and can deliver the promised performance for the vehicle. A customer may love the look of a set of custom wheels. But if they don’t meet safety standards or weigh so much that they hurt the car’s fuel economy, the company can’t sell them. Enter AI: the latest tools for generative product design help designers automatically create parts optimized for weight and strength. These tools can explore a design space rapidly while still meeting the constraints of manufacturing. Meanwhile, small-batch manufacturing is increasing, and automation is making it easier to move from design requirements to a custom manufactured part. Given these combined advances, generative design tools will soon power business-to-business and even consumer interactions.
We built a demo of this approach for the automotive space. A customer purchasing a vehicle uses a website to easily design and order custom car wheels. Building on the base capabilities of generative design to explore a design space, this demo leverages data from the customer’s current car or a test drive to incorporate their driving preferences into the design considerations. But it also enforces their stylistic preferences as well as aesthetic guidelines from the carmaker’s brand. The end result: a large set of completely custom wheel options that match the user’s driving and style preferences, while still staying true to the design of the brand. With generative design, companies can bring a high-end customization experience to all customers.
Across all of our efforts, we’re pushing the boundaries of advanced technologies to enable the future of design. Immersive experiences, intelligent tools, and new level of collaboration capabilities will all play a role. Watch this space, and reach out to us to learn more! For more information about the platforms from this piece, contact Alpana Dubey, Nick Akiona, and Mike Kuniavsky.