Explore three ways AI already impacts the design and engineering fields, from generative design to advanced robotics.
This article is based on a Raconteur report about the democratization of product design. You can download your copy of the full report here.
Recent advancements in computational power, cloud connectivity, and neural network complexity have ushered in a new era called “The Augmented Age.” Artificial intelligence (AI) tools already augment many aspects of our daily routines. In the next five years, a vast majority of product design companies also anticipate relying on them heavily. Maurice Conti, who coined the phrase, describes a world in which AI augments the human ability to design, build, and employ products at every step of the design process.
This article explores three ways AI already impacts the design and engineering fields.
Is generative design AI?
Generative design harnesses the power of artificial intelligence (AI) to generate design iterations based on constraints identified by the designer or engineer. Typically, generative design is used to reduce weight, reduce manufacturing costs, and optimize part performance. Generative design uses AI to some extent, but it is just one step in the process. Humans are required at the beginning of the process to define constraints and at the end of the process as the final decision-maker. After all, someone needs to sift through the list of design iterations generative design creates!
Generative design for ideation
Ideation is the first and most striking stage of AI-enhanced product development. Even agile design teams typically iterate less than a hundred concepts and prototypes throughout a project.
Today, using generative design offers the ability to explore thousands of possible solutions across the entire design space. “Rather than starting with a drawing, the designer starts with the parameters required for the end product,” Peter Champneys, a research engineer at Autodesk describes. “The system then generates thousands of potential solutions in just a few hours and identifies the top few options that best fit the requirements.”
Not every AI-generated design is a viable solution, however. AI design results can be impressive, yielding lighter, stronger, and more manufacturable products than ever before. “The human contribution is to then evaluate the outputs and determine the best performing, most cost effective and visually attractive solution.”
AI & advanced robotics for development
An often overlooked application of AI in the product design cycle is using robotics for prototyping and building. Additive manufacturing, CNC machining, automated welding, and many other tools rely heavily on AI to achieve functionality and precision beyond human capability. Advanced robotics can do much more than mundane, repetitive tasks. Robots can even utilize sensory inputs like vision, touch, temperature, and sound.
The Hive Project
An excellent example of the synergy between humans and artificially intelligent robots is the Hive Project. This was a collaboration between Autodesk and The Institute for Computational Design (ICD) at the University of Stuttgart. The goal was to construct a specialized three-dimensional surface using bamboo and string.
Machines and robots using AI were employed as high-level construction managers, precision machinists, and string assembly technicians. In contrast, humans selected and manipulated the highly irregular bamboo pieces. Together, they created a structure that couldn’t have been built without human and AI collaboration.
The final piece of the puzzle bringing us into the augmented age is the closing of the design loop. This means feeding product lifecycle data back into the iteration process. Closing the loop can happen in two ways. The first is the aggregation and analysis of sensor data from within the product. And the second is customer feedback from interacting with and using the product. Between the widespread adoption of wireless sensor nodes and cloud-based computation and storage — that’s a lot of aggregated data. In fact, it’s often too much for humans to interpret.
Here, AI can be relied upon to extract meaningful data trends and inflection points in how a product performs in real-world applications. This data can then be fed back into the generative design process with the help of a human designer to update parameters and optimize future changes in the product.
CityBot by EDAG Group
One example of a closed-loop, generative design is CityBot. Citybot is a modular, configurable, multi-functional mobility platform from EDAG engineering. It’s an adaptable mobility robot that can reconfigure into different form factors depending on the application — a forklift in a warehouse to a trash collector in a public park. Customer requirement specifications and usage data from CityBots are fed into a generative design loop to realize an optimized product.
As Sebastian Flügel, Project Leader at EDAG Group describes it: “The big change [from the old process] is that generative design creates the geometry automatically, and with Fusion 360, we can automatically recalculate these products, and we can create new solutions based on the changed conditions.”
“With all the sensors embedded in our parts, we know precisely if the part is stressed, which parts need to be improved, lighter or stronger for each function,” Johannes Barckmann, Global Design Manager at EDAG Engineering, explains.
This automated and continuous optimization type has never been seen in traditional manufacturing environments. The result is unprecedented flexibility and customer satisfaction.
The reality of AI adoption
The widespread adoption of generative design, advanced robotics, and closed-loop engineering will require more than just technological advancement. A widespread cultural shift will need to occur. Engineers and designers must leverage the value of AI in their day-to-day thinking to realize the full potential of the augmented age.
To learn more about the future of making, visit Autodesk’s The Future of Product Design and Manufacturing.
Ready to harness the power of generative design in Autodesk Fusion 360?