A Hierarchical Task Network planner utilizing LLMs like OpenAI's GPT-4 to create complex plans from natural language that can be converted into an executable form.

Lightweight C++/Python Hierarchical Task Network (HTN) AI Engine

Hierarchical Task Network example in Godot

6 DOF arm manipulation, utilizing knowledge-based reasoning, BDI and HTN planning

Automatically creates a solution to a randomly-generated maze using Pyhop, a Hierarchical Task Network Planner created by Dana Nau based on the SHOP algorithm. Built for artificial intelligence research.

Automatic Generation of HTN Domains from VGDL Videogame Descriptions

A 3D first-person capture the flag game made with Unity in which the player must capture a treasure chest guarded by a Hierarchical Task Network (HTN) based AI.

Model that utilizes Pyhop, an A.I. planner based on the JSHOP2 algorithm, to determine whether a person needs assistance based on task progress. Currently integrated into the Computational Model of Assistance. Done during the Summer of 2019 as Research Assistant to Jason R. Wilson, post-doctorate researcher at Northwestern University.

A fighting game AI: KeepAwayBot, implemented in Java using Hierarchical Task Networks and the Upper Confidence Bounds Algorithm

HTN planning including resource and temporal constraints with SHOP3. Repository for my Bachelor Thesis @FHAachen.

Domain model for Hierarchical Task Network Planning (HTN) implemented in Unity C#