This project is based on the article: Watching and Acting Together: Concurrent Plan Recognition and Adaptation for Human-Robot Teams by Steven J. Levine and Brian C. Williams (2018) from MIT Computer Science and Artificial Intelligence Laboratory.
In recent years, the field of AI (Artificial Intelligence) and especially the use of robots, has been developing rapidly, among other things, for the purpose of performing work together with humans in real time.
Nowadays, there is a very large demand for humans and robots to work alongside each other, in order to perform various tasks. In order to succeed, robots must be able to recognize the human intent and adapt accordingly.
PIKE presents a unified approach to these two problems – Recognition and Adaptation. PIKE is an executive for human-robot teams, that allows the robot to continuously and concurrently reason about what a human is doing as execution proceeds, as well as adapt appropriately.
PIKE’s architecture contains two parts – offline and online:
- Offline – In this part, PIKE analyses all possible choices that a human can take. PIKE uses optimization algorithms in order to find the best possible set of choices for the robot and assist the human in the most efficient way. Eventually, the robot prepares databases of all the best possible choices.
- Online – In this part, PIKE uses the pre-analyzed databases scenarios of human choice, along with his real-time decisions. Then, PIKE chooses the best course of action for the robot according to the specific situation.
In our project, using Python, we implemented a significant part of PIKE’s offline algorithms.