Evolutions of robotics require basic machinaries of how human interacts with each other. Some may argue these basic machinaries are uniquely belonged to human beings. We believe we can design very different machines based on these machinaries. Can a robot like or dislike a person? How can a robot build a relationship with its friends? Can a robot think based on its relationship with human? Can a robot learn in an irreversible manner? Shybot is here to explore these questions. Shybot is an emotion-driven robot like a shy creature which recognizes friends or strangers and interacts with them differently. We envision the Shybot project can be further developed into the software core for emotion robotics. According to Asimov's foundamental assumptions for robots, we believe the robotics design has to be intricically coupled with human-machine interaction.
- A robot may not injure a human being or, through inaction, allow a human being to come to harm.
- A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
- A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Shybot was initiated as a research project from the Autism Theory and Technology class (co-taught by Rosalind Picard, Cynthia Breazeal, and Sherry Turkle) at MIT Media Lab in 2007. Many of Shybot's features are inspired by The Uncanny by Sigmund Freud (1919) and Computing Machinery and Intelligence by A. Turing (1950).
-
Peer-to-peer face learning. It would be more efficient to accumulate the "learning" part so that this network of Shybots could keep improving themselves.
-
Companion. The plan is to re-build it using some existing platform (e.g., Romo, SparkFun robotic car, or Raspberry Pi with camera, sensors, and motors). This repository contains Shybot's work-in-progress prototype code based on Romo. For making it work as a consumer toy, we'll need durable design and robust sensors for basic safety concerns. Adding human detector (and danger detector) using non-contact temperature sensors (MLX90614 or MLX90620). This part was paused because it's also too much fun to build its physical part. We'd like to focus on the software now.
Lee, C.H., Kim, K., Breazeal, C., Picard, R.W. Shybot: Friend-Stranger Interaction for Children Living with Autism, Work-In-Progress in the Extended Abstract of Computer-Human Interaction 2008, April 5-10, 2008, Florence, Italy
- All codes are under MIT license.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent