Generative design: leveraging artificial intelligence for design optimisation

vice president of Solutions Consulting at PTC, explains.

Generative design is transforming the world of design engineering and has the potential to deliver on some of the broader challenges around sustainability and the circular economy. In its simplest form it is a technology in which 3D models are created by computer software allowing greater optimisation for purpose and reducing the design cycle.

The basic aim of generative design is to allow engineers to spend more time engineering and less time iterating on design alternatives – which encourages creativity from the outset and enables engineers to realise their ideas in ways that had not previously been possible. The software driven solutions often go beyond the limit of their personal experience by leveraging powerful artificial intelligence (AI) capabilities that guide the next generation of high-performance product designs.

The Engineer talks: watch our video interview with Paul on generative design

The Engineer talks: watch our video interview with Paul about industrial applications of augmented reality

In practice, a user sets up a model containing factors such as traditional loads, constraints and materials together with the available geometry envelope and associated manufacturing process options. The generative technology then delivers a series of optimised designs that meet those requirements. The designer is still fundamental to the process and needs to factor in parameters such as physical constraints and the availability of materials. In fact, the more astutely the engineer defines these criteria the better the outcomes will be. It is all about helping engineers become more productive by allowing them to explore more design iterations and more ideas than traditionally would be possible for them in the conventional design cycle.

Built on foundational technology

Traditional topology optimisation and generative design have often been used interchangeably in the market and this has caused some confusion.   Topology tools typically start with an existing design and optimise it  based on functional objectives such as maximum material stress or deflection. Generative design includes the same material stress and deflection goals but starts with an available space envelope and the manufacturing options to arrive at a range of solutions which fit the functional and production requirements. It uses topology optimisation, often performed in powerful cloud computing environments to explore all the possible permutations of a solution, continually iterating, and learning from the process. And it’s iterating far more times than traditionally possible in a typical design department where there are always time or resource constraints.

Working in a single interface

For PTC, the journey began back in November 2018 when we acquired a technology company called Frustrum and since then we have worked to develop that original code base. It is now fully embedded inside the Creo environment, which means there is no need to export geometry or leave the Creo interface.

Within this solution there are several options for engineers. The first is Generative Topology Optimisation, which runs and delivers generative design on the hardware that the engineer is working on, the workstation or laptop. In this scenario the power is right there in the software. We also offer a design extension for generative design that makes use of our cloud capability, which is part of a broader cloud strategy within PTC.  Within the cloud option, you are not just looking at one design alternative, but we can deliver many options depending on the variety of goals.

Saving time and weight

With the growing global interest in space exploration, a topical example of generative design comes from a US based company who design and manufacture human life support systems for space suits – used by any astronaut during a spacewalk. Its full name is the Exploration Portable Life Support System, and astronauts on the International Space Station rely on it to keep them safe whenever they leave the confines of the station. The engineering team wanted to achieve the optimal balance of performance and safety, which is obviously critical in that environment.  And to achieve this they used generative design to explore hundreds of different design combinations of materials, manufacturing processes with the desired safety factors and stress levels within the materials, so they could achieve an optimised solution.

There is an interesting quote from one of their engineers that many of us can relate to.  “As an engineer I like right angles, flat surfaces, and rounded dimensions,” he said. But of course, generative design often comes back with more organic solutions to the problem and therefore as an engineer he said, “I have to accept that I might not have the perfect solution and that generative offers something better than we would normally come up with as engineers.”

The solution from the generative design process delivered a 50 per cent mass reduction that helps the astronaut in terms of their manoeuvrability while they are in space. Not only that, they estimated they were able to shorten the design cycle by 20 per cent.

Driving sustainability with generative design

Generative design also has an important future role to play in sustainability and the circular economy, particularly in enabling engineers to understand the carbon debt that they are building in when selecting materials and the associated manufacturing process.  One thing that we are looking at is the ability for us to embed a specific goal in the generative design process such as enabling the minimum carbon debt for a component. That is a function of the material, the manufacturing process – whether it is additive or subtractive. It even comes down to the carbon footprint of the material supplier. If, for example, the steel you’re using was created in an electric arc furnace, you know that is better for the environment than if it was generated in a traditional blast furnace. By expanding the objectives and goals through generative design we can include environmental and sustainability topics.

With all these potential benefits, generative design is becoming more and more appealing to manufacturing companies as they seek ways to optimise new products for improved manufacturability, reduce material costs and decrease cycle times. I think we’re just starting to see the full potential for this exciting new technology.