Edge 393: Understanding Planning Techniques in Autonomous Agents
A taxonomy of planning in autonomous agents, the ADaPT planning method and the XLANG framework.
In this Issue:
We start diving into planning methods in autonomous agents!
Review the research behind Allen AI’s ADaPT planning method.
Provide an introduction to the XLANG framework for building autonomous agents.
Let’s go!
💡 ML Concept of the Day: Understanding Planning in Autonomous Agents
Planning is, arguably, the most important capability of autonomous agents. By planning, we refer to the ability of taking an abstract goal as input, decomposing it into specific tasks and execute those tasks. Every autonomous agent framework incorporates some planning capabilities even though there hasn’t been a formal definition of the space. Obviously, there are many ways to implement planning in autonomous agents. A recent paper from AI researchers from the University of Science and Technology of China, Hefei, China outlines a taxonomy for planning methods in autonomous agents that we find quite helpful.
In their paper, the researchers identified five forms of planning in autonomous agents: