Autonomous Design

Test Bed Vision:

How do we optimise complex things?  The convergence of new materials and manufacturing processes (including 3D printing) are opening up a new frontier in the provision of bespoke, high-performance physical artefacts encompassing a myriad of material-geometry combinations, that are set to revolutionise a range of impact areas in industry.

Photograph of a T-shirt print, the robot thinker.

Photograph of a T-shirt print, the robot thinker.

Successfully navigating this frontier requires a new breed of design algorithms, to efficiently traverse these huge, complex, multi-dimensional spaces, either assisting or replacing human designers and engineers. Efficient reality-sampling modelling techniques allow us to blur the lines between the real and the virtual; offloading the bulk of design iterations onto high-performance compute clusters whilst maintaining veracity.

We particularly focused outputs in the area of robotics; a key area for the economic health of Australia.  Combining materials, manufacture, modelling, and design allowed us to conceive a new type of robot underpinned by smart physical embodiment, able to perform previously-impossible tasks in industrial settings, as well as allowing deployments from deep sea to deep space and completing complex long-term missions in the harshest environments Australia has to offer.

This test bed was led by David Howard from CSIRO’s Data61 , and was a collaboration with staff from CSIRO’s Energy and Manufacturing.