I have been developing a drawing robot within the context of the Arts with three core motivations. First, to challenge the standard approach to designing robotic systems. Second, to investigate how we as people draw and perceive marks. Third, as an initial experiment in Robot Poetics - I am interested in how the very pursuit of making robots reflects our humanity, and what we can therefore learn from reading (perceiving) robotic artefacts and artwork. I have written up these ideas in an Open Access academic publication available here. There is also an early interview with The Bristol Magazine available here.
However, asking questions of humanity makes things a little complicated. For instance, how is it that we recognise 'good' artwork? There is no formal solution to creativity or aesthetics. Within science and engineering it is a little easier to assess whether a robot is good enough - we can ask if the robot performs its task well. In comparison, the expectations of a robot within art are broarder, and it will be perceived by people who bring with them diverse backgrounds and influences. Therefore I argue that Robot Poetics is to pursue robotics as more than a tool (transcending pragmatism) and to intentionally aim for cultural significance.
Of course, robots are already well within our cultural sphere, appearing as the characters Data, R2-D2, The Terminator, as well as prompting debate on such topics as whether real robots should be gendered, etc. Given the prominence of robots in culture, Robot Poetics is really about asking what an ideal form of a robot as a cultural artefact would be. The idea of Robot Poetics can be applied in hindsight to all the robots that exist (fictional and real) and discussed as to how poetic they seem to be. However, to build robot poetically (a forward process), I think it is important to carefully consider how every attribute of the design, fabrication and existence of the robot will contribute to its cultural significance.
Because robots are popular and spark the imagination, and due to anthropomorphism, I felt it important to design my drawing robot to be as simple as possible (but not simpler). In effect, I wanted to reduce any aspect of theatre to a minimum - I don't want my audience to feel tricked. This is a nuance to the idea of Robot Poetics: whilst a person could be deceived into thinking a robot is human-like or has human-qualities (a simulation, or theatre), poetics has value when it reveals something different, unexpected or unknown. We know that a robot is not a person, fundamentally (i.e. if it approaches personhood it is no longer really a robot). Therefore, it is authentic to instead ask in what way and how a robot can exist in juxtaposition to a person.
What is sought in Robot Poetics is a shift of perspective on what we think we already know. For example, in written or spoken poetry, this shifting-effect is can be achieved with metre. Metre can make words seem to have melody or rhythm, which enhances the effect of the words beyond their simpler pragmatic reading as information. With robotics as an artistic medium, there is the potential for a vast range of poetic devices (phenomenon to alter perception and understanding), from the materials of the robot, a robots operation, up to how the robot is received by an audience, and beyond.
It may be the case that instances of Robot Poetics do take quite literal forms, such humanoid robots. There is still something to be learnt in a literal approach. It is also the case that there will be a spectrum between unsucessful to sucessful Robot Poetics. However, to learn something about humanity, what is aspired to, investigated and analysed, is the ideal form of Robot Poetics. This is really the central question, what is an ideal form of Robot Poetics? This question then generates lots of other questions to investigate as stepping stones on that journey. The drawing robot is one such stepping stone.