Categories
_

Science makes art. But could art save the Australian manufacturing industry?

The “hand-made” nature of artistic works has been highly valued by humans over thousands of years. But digital capability is changing art – not just how art is designed; also how it is made.
Now we’re at the point where the art and design industry in Australia is demanding “mass customisation” of artworks. Some companies have started to address this using the latest generation of robotics technologies – but making the technology work in the right way needs input from creative expertise.
Done right, this mashup of creativity with technology could strengthen manufacturing capability in Australia.

Gap between design and production

There is a tremendous gap between the ease of digital design and manufacture of bespoke objects.
When computer assisted design rapidly evolved in the late 1990s, it meant that previously impossible to conceive ideas could find a form on the computer screen.
But the act of actually creating computer-designed works can be more difficult – and costly. Architect firm Gehry Partners used digital design software to design a “crumpled mirrored” staircase for the University of Technology Sydney. But when creative company UAP manufactured that staircase (shown in the photo above, and animation below), they had to employ a thousand year old technique of meticulously hand beating every surface until it matched the shape of the computer model.

The ‘crumpled mirrored staircase’.

Reproducing or re-sizing works of art can also present manufacturing challenges. Originally, artisans would carefully measure the original object and then hand craft the copies, sometimes adjusting the scale. Now, modern scanning technology can create highly accurate computer models of such objects – but the same problem of how to manufacture the new objects presents itself.
The technology to take a digital design into a mechanical fabrication process exists, but it is normally too costly for one-off pieces and is reserved for mass production.
This is where robots come into play.

Robots that see

For a robot to make something where the starting form and desired final shape are not fixed – that’s complex.
Traditionally, robots have been used for manufacturing tasks where the shape of the object being worked on is very well understood. For example, robots can be used to remove the excess metal (a process known as “fettling”) after metal casting of car engine blocks. A robot can be programmed to do this as the desired final shape of the engine block is known: without visual information, the robot can move the engine block over a grinder to remove any excess metal.

Robotic fettling of a known object.

But many of the objects created by artists do not have a detailed computer model for the robot to work from. Also, works of art are typically not uniform or predictable in shape. So any robot working on a piece of art will first need to see it from all angles, and accurately discover its shape.
The technology to see, or scan objects exists now. In fact you may have it on the smart phone you own right now.

3D cameras that scan objects in detail are already on the latest smart phones.

But the next challenge is determining how to work on the object: could a robot transform an object it sees into one that is desired (a piece of art)? We’re not too far off this goal.

Robots might create jobs

Many people fear job losses associated with introducing robots into production facilities. However the number of jobs can actually increase when robots are used in mass customisation.
In our own discussions with UAP – the company that made the Gehry Partners staircase – they tell us that since adopting robotic technology, staff numbers have grown at a rate of six new appointments for every piece of new robotic machinery purchased. Existing staff are working in new technologies; for example, pattern-makers are using their expert sculpting skills in virtual reality and sending these digital works direct to robotic manufacture.
The products UAP are making with robots include artworks like Poll (by Emily Floyd), and architectural facades that will soon be installed on busy city streets in Australia.
This sort of mass customisation manufacturing may also be suited to products such as customised stents for arteries, or even production and preparation of better looking fruit and vegetables for niche food markets like airlines. The workforce may grow as a result.

Let’s invest in creative skills

Design is a fundamental creative manufacturing capability.
Currently in Australia, government manufacturing policies and investment programs have a firm focus on supporting science and technology companies such as aerospace companies printing jet engines, or biomedical science entities growing parts of the human anatomy. And while the strategic importance of robotics to our manufacturing future is well established and funded, this is not the case for design.
Creative capabilities in art and design firms should be more strategically included in this investment.

Recent data shows digital creative services are growing at nearly three times the rate of the overall workforce and attracting 30% above the average Australian wage.
The ConversationRight now, Australian governments should be targeting the innovation capability of the creative industries, and expanding the value art and design already add to Australia’s manufacturing industry.
Cori Stewart, Director, Business Development and Associate Professor Creative Industries, Queensland University of Technology and Jonathan Roberts, Professor in Robotics, Queensland University of Technology
This article was originally published on The Conversation. Read the original article.

Categories
_ News

Smartgeometry Workshop and Conference

Research Fellow Dr Muge Belek Fialho Teixeira was selected to participate in a workshop at the Smartgeometry workshop hosted by the University of Toronto earlier this year. In this post, Muge reflects on the workshop and conference.

Smartgeometry was founded in 2001 and is now a biannual event.  It starts with four days of themed workshops followed by a two day conference. Smart Geometry (SG) workshops and conferences have been influential to many disciplines including architecture, design, engineering and mathematics. Originating as a collaboration between industry, researchers and academics, SG has always been a platform where innovative ideas become a reality, informing the potential needs of the disciplines towards a better future.
The workshops are called clusters and are organised around open calls coordinated by ‘cluster champions.’ Cluster champions are collaborative teams from academia and practice who get together to prepare a proposal, or a response, to a specific theme. SG’s open call encourages researchers, academics and industry to discuss possible research questions around the proposed theme and a research avenue, via a project. By working on this project, researchers and practitioners from industry and universities have a chance to see how these technologies can be applied. Participants for each of the clusters applied for a position via open calls with cluster briefs defined by cluster champions. Participants were selected, from a competitive, international pool of applicants, based on their background, research expertise and current interests.

The conference, which took place after the workshops furthered discussions around the workshop themes informed by different perspectives from multidisciplinary invited keynote speakers. The conference was curated in a way that would feed back into the outcomes from the workshops. In that manner, it was a dynamic conference, where the keynote speakers build on the work produced by the clusters and open up new agendas for future speculations. The conference was followed by Q&A sessions that allowed the workshop participants to engage with the keynote speakers openly. These exchanges also provided opportunities for future collaborations.
The University of Toronto hosted Smatgeometry under the theme “Machine Minds”, which revolved around machine learning and AI (Artificial Intelligence). Current discussions on machine learning and AI, generally consist of depressing scenarios of humans coming to an end or humans losing their jobs. Websites like “Will robots take my job?” are opening up discussions about how we should give away our passions for our professions. As a trending topic for many disciplines, SG focused on how machine learning and AI can be utilised for design and what could be some other positive and constructive ways of approaching this topic. The clusters explored the applicable areas of Machine Learning and AI, whereas the keynote speakers of the conference tried to create an understanding of what is machine learning and AI and its impact on our society, as well as the methods they use them in their practice.
The clusters at SG were:
–          Smart materials (Fibrous timber joints, Materials as probes)
–          Smart geometries (AI strategies for space frame design, Mind ex-machine)
–          Smart fabrication methodologies (Soft Office)
–          Smart and innovative ways of perceiving the environment (Behavioural Enviro[NN]ments, Data Mining the City, Fresh Eyes, Inside the Black Box, Sound and Signal)
All of them used cutting-edge technologies and customized software to define geometries. These technologies included interactive tables, VR headsets, industrial robots, mobile robots, CNC routers, sensors, microphones, and many more. One of the most dominant software platforms used by clusters was Rhino with the Grasshopper plug in, as a unifying platform, but there was also other software such as Unity, Processing, Arduino, Python, or custom build software for the clusters. More information on each of the clusters can be found here.
Highlights from conference discussions were;
–          what is AI and machine learning,
–          how AI and machine learning will affect the future of societies and how we can get prepared,
–          collecting, interpreting and managing data,
–          natural intelligence versus digital intelligence,
–          machine learning versus human learning,
–          robotics and advanced manufacturing,
–          interactive installations,
–          complex geometries.
The schedule and the keynote speakers can be found here.
As part of the SG2018 there was also a trip to see the new workplace of Autodesk Toronto. Autodesk has been a close collaborator of SG as a sponsor and providing know-how, keynote speakers, cluster champions and event participants. The new Autodesk workplace has been designed using generative algorithms and has a research centre for exploring new technologies. One of the clusters (Mind ex Machina) took place in this research centre, using two UR10 collaborative robotic arms with custom build open source software for SG18. It seems Autodesk has started to take a pioneering role in research by collaborating with research institutions, researchers and companies through these research centres. With artist-in-residency programs, they are opening up their facilities globally to makers and curious minds. A list of Autodesk research centres can be found here.
Looking forward to the future, next Smartgeometry will take place at Carnegie Mellon University in Pittsburgh, USA, 2020 with another challenging theme!