Robots and Magic Cards

Article: Magic Cards: A Paper Tag Interface for Implicit Robot Control

Author(s): Shengdong Zhao, Koichi Nakamura, Kentaro Ishii, Takeo Igarashi

Summary: To begin this article, the authors describe the fast moving progress of robots out of science fiction into the real world, and provide examples of robots that are already popular in society, such as the Roomba and the Scooba, which are two robots used to clean surfaces. The authors also describe various stories and folktales that talk of gnomes or magical beings doing tedious chores for them, like housework, while they are asleep or away from home. The authors then describe an idea they have come up with to abstract away the layer of complexity that exists when users are forced into interacting with robots in order to instruct them to perform tasks. The idea is to remove the person instructing the robot via button inputs and move it to a "Magic Card" system. The Magic Card system works by having various types of cards used that can be placed in areas or on objects, that instruct robots to do various things at various times, or to avoid certain objects while accomplishing other tasks. The user, which in this case is a typical male or female doing housework, would place a card such as "clean this room" and "at noon" in order to tell the cleaning robot to clean a certain area at noon. These cards could eventually be memorized in a certain order to allow repeated tasks to become simplified. However, due to the current state of robotics, the article mainly only dives into cards that can be read on the ground, and performed with robots that are currently capable of doing ground-level chores, like vacuuming. The authors did a study in which they paid participants to give them feedback on their system, finding that there was very little confusion with the cards, and almost all of the participants were excited about the idea of the simplicity of having robots do chores while they were away simply with cards.

Shaun's Opinion: Kudos to the authors for their very creative method of abstracting the layer of complexity of robotics away from the end user, especially the elderly or disabled. This idea could save quite a bit of time for most people, if they were able to get things done while away simply by purchasing some robots. The card system is a very good one, with simplified tasks in an easy human readable form such as "clean this room at noon". However, the tag system itself seems to create a bit of tediousness in its implementation. In order to perform a task, the human user must create all of the tags and place them in the appropriate spots. For instance, the authors describe a system where the users could place tags on objects they wanted to be moved to a single location. The system, in this case, requires the user to go to each one of the items and "tag it" to be picked up, when the user could probably have just done it themselves in the same amount of time. Not to mention leaving these tags all over the place in your home seems to create a mess of its own. However, the possibilities that arise from an idea such as this have great commercial potential, especially if the interaction between users and robots could be simplified to make it available to a wide range of people, technical-savvy or not.