Generative Design and 3D Printing

–And its relationship to healthcare 3D Printing and Biofabrication

In the world of 3D printed anatomical modeling, the focus is to replicate human bodies (“digital twin”) and stay as accurate as possible to what nature’s design is, pathology or healthy.

But what if the human body is just one of the countless design possibilities that would work equally well with the environment? And if the human body is an assembly of parts, then each body is also a composition of many more magnitude of such design possibilities.

The term generative design surfaced again recently on the cover of Science magazine (Figure 1)[7], which demonstrates a bio-printed artificial lung component, simulating the form, vascularity, and functionality of a human alveolar using generative design by the design firm Nervous System.

Figure 1.

Here is a video as well.



Understanding the concepts behind the generative design is not only required to fully explore the potential of 3D printing in the life sciences, but it is also a mind-blowing experience and raises metaphysical questions such as: Do you think our existence is accidental? Or, do you believe the world is simply one choice out of a billion other equally good possibilities? Can we design bodies even better than what we have?  I will provide an introduction to this subject, a mind-blowing new world of design, and I hope this can stimulate interesting discussions and creativity. This is the first but definitely not the last bog on this subject from me.

Generative Design – The Concept

Like “additive manufacturing”, the term “generative design” suggests an abstract process rather than a specific machine or software. However, it’s only when “3D printers” and “generative design software” became available, these concepts became exponentially impactful in our lives.

In Wikipedia [1], generative design is defined as “an iterative design process that involves a program that will generate a certain number of outputs that meet certain constraints, and a designer that will fine-tune the feasible region by changing minimal and maximal values of an interview in which a variable of the program meets the set of constraints, in order to reduce or augment the number of outputs to choose from. The program doesn’t need to run on a machine like a digital computer, and it can be run by a human for example with pen and paper. The designer doesn’t need to be a human, it can be a test program in a testing environment or artificial intelligence, for example, a generative adversarial network. The designer learns to refine the program (usually involving algorithms) with each iteration as his design goals become better defined over time.”

Here is a simple illustration of the design process from Wikipedia (figure):

Here are two diagrams comparing the traditional design process with the current generative design process using AI/ML.[8]

Source: CMIDATA
Source: CMIDATA

Generative design software like Autodesk Dreamcatcher uses machine learning/ artificial intelligence to mimic nature’s evolutionary approach to design. It is a “generative” process because the computer generates entirely new designs that designers never seen or even thought of before. Once a designer has a well-defined idea, he/she can input design parameters (such as functional requirements, performance criteria, size, materials, weight, manufacturing methods, and even cost constraints) into the software, and will be able to explore hundreds and thousands of possible options. The computer will produce thousands of new designs, conduct simulations, generate performance data, and continue to refine its search according to the designer’s design goals. The designer can evaluate the generated solutions and their performance data in real-time, adjust design parameters and goals, accelerating the creative process. [4,5]

Here is an example of what the generative design process looks like:

Technical Benefits of Generative Design

  • A generative design system can leverage artificial intelligence to generate a much larger number of possible solutions with given design parameters and constraints that are more than any human designer can do. In a way, it is fulfilling the “augmented” designer goal mentioned in this TedTalk by Maurice Conti. This advantage of the generative design system is profound, as new bio-printed lung alveolar tree by Nervous System as an example. Perhaps, new forms of human organs with equal or better performance can be created in this process? This will leave many to ponder the metaphysical aspect of a generative design system.
  • Generative design software bridges the gap between the cognitive/virtual world with the physical world by accounting for manufacturability. In other words, there is no point in a great looking design on computers without figuring out the best manufacturing strategy for the design. Since the generative design system tends to produce products that have complex geometries typically impossible for the conventional manufacturing process, the system can provide manufacturing inside accounting for different manufacturing options (3D printing, CNC, casting, etc.).
  • Generative design and 3D Printing (or Bioprinting) are a match-made-in-heaven. One of the major advantages of additive manufacturing is “complexity for free”, moving the manufacturing process from an “assembly” to “part consolidation” without additional cost. Direct benefits of this include saving materials (improve sustainability), decreasing part’s weight (topology optimization), saving time by decreasing labor cost from the traditional assembly process, saving money by decreased capital investment in traditional manufacturing setup (e.g. injection mold).  On the other hand, many designs from the generative design system with great performance data could not be manufactured in any other way but 3D printing.

Examples of Generative Design in Healthcare

To give our readers an idea of what generative design can do in healthcare (in addition to the bioprinted lung alveolar above), here are a few good examples.

The first example is orthosis by MHOX and CRP Group, using an integrated workflow of 3D scanning, generative design of 3D models, and 3D printing to achieve mass customization of a generative orthosis. The generative design process takes into consideration specific performance needs, mechanical properties, aesthetic values, material properties, and more. The hand orthosis is intended for post-surgical rehabilitation, aiming to limit any movement. The porosity allows the hand to stay ventilated and washable. [3]  

The second example is a wheelchair designed by Rachael Wellach, founder and CEO of Disrupt Disability, and Steve Cox, 3D Tech Consultant, AMFORI Consulting. [6] They pair aimed at creating a modular wheelchair that can be aesthetically pleasing, affordable, low cost, and personalized.  

Conclusion:

The concept and tools surrounding generative design are still very new, and it is already “augmenting” a small group of designers and engineers who can now create products that no one could have never created before. If human body is only one design choice out of a million other possibilities, perhaps generative design process is a way to find and create the ultimate optimized body parts in a form that we have never seen or experienced before, instead of just a foreign replacement, and with the newly available manufacturing process such as bio-fabrication and additive manufacturing.

References:

  1. https://en.wikipedia.org/wiki/Generative_design
  2. The New Age of Highly Efficient Products Made with Generative Design
  3. http://mhoxdesign.com/generative_orthoses-en.html
  4. What Generative Design Is and Why It’s the Future of Manufacturing
  5. https://autodeskresearch.com/projects/dreamcatcher
  6. How generative design and 3D printing are making customised, lighter wheelchairs a reality
  7. Science (Science  03 May 2019: Vol. 364, Issue 6439, pp. 458-464 DOI: 10.1126/science.aav9750).
  8.  The Next Wave of Intelligent Automation by Harvard Business Review

Related Articles:

When Artificial Intelligence Meets 3D Printing

Reimbursements for 3D Printed Anatomical Models and Surgical Guides – Decoding the CPT Codes

Decentralized Healthcare — Part I. Cooler than Bitcoins, But What Is It?

3D Bioprinting: Chiasm of Art, Design, Science, Technology, Evolution

Cancer: What 3D Printing (Bioprinting) Can do For Oncological Care

Cybersecurity for 3D Printed Medical Devices, Less Crazy and More Useful Than Bitcoins

Comments