The objective of this project is to create a robotic system with the Sawyer robot arm that draws pictures with a pen. In particular, we set the goal to reproduce a portrait image of a human face, which should be recognizable without viewing the original image. We set the expansion to multi-color images as a stretch goal.
The problem is interesting because it requires high accuracy end-effector pose control in a plane. In other words, the accuracy requirement is limited to a plane, but also means that traditional control methods such as inverse kinematics are no longer practical. Furthermore, the project requires multiple representations of the same data, with minimal deterioration, and conservation of significant features.
Robotics has the potential to revolutionize a multitude of industries by replacing humans in repetitive, hazardous, and dangerous work, allowing people to focus on more creative and stimulating work. Furthermore, with the continuous research effort in fields such as surgical robotics, robotics is already saving lives directly and will save countless more in future years. In short, robotics will without a doubt have a net positive effect on the world and our lives.
Despite this, however, public perception of robotics is excessively negative, in part due to popular culture films such as The Terminator (1984), 2001: A Space Odyssey (1968) and Blade Runner (1982), where robots are represented as relentless killing machines with no humanity. Furthermore, robotics is seen as an intensely technical field, discouraging young students who favour creativity and expression.
We wish to change this for the better, by showing a human side of robotics, namely art, which explores the many aspects of human experience. The subjects chosen for this art project reflect intensely human ideas such as the abstraction of faces, cultural archetypes, surrealism, and impressionistic modernism. Furthermore, while the images are intended as reproductions of human creations, the style of the artworks is serendipitously characterized by the robot arm implementation through Ada.
Robotized artwork could further open up artwork creation to people who are physically unable to draw. Our robot arm can switch between pens without needing human assistance.
Furthermore, digital artwork that came from a printer does not look the same as art that was drawn with a pen or painted by hand. If an artist wants to mass produce their artwork, they could record their pen strokes and have a robot repeat them to create multiple copies.
This can increase the economic opportunity for artists who wish to share the in-person experience of their work with more people rather than sharing it online where differences in how different screens display the same 'color' can cause the same artwork to be perceived differently to different people rather than conveying the feelings the artist intentioned.
Someone could industrialize this by running a website and art factory where customers can submit pictures they want detailed pen sketches of, automatically sketching it, and then shipping them their sketch. This would reduce the cost of self portraits or artistic family photos, making personalized artwork widely available to the average consumer.
Artists who make artwork people don't like will lose out because people wouldn't have to buy from them anymore and can choose to buy copies of high quality artwork or personalized automated artwork. If the robots are run using clean electricity, like that generated from nuclear energy, there will be little impact on the environment.
Beyond art, the notion of a redundant robot system with accurate manipulation capabilities in a given plane may be extended to applications such as pick and place on tables/floors, sliding on flat surfaces, and surface characterisation through sweeping.