Implicit Communication in Human-Robot Collaborative Transport

Requirements

Core

• Python 3.10
• uv package manager
• ~2 GB disk space for Python packages

Extra Dependencies for Deployment and Data Analysis

• ROS 2 Humble (Ubuntu 22.04)
• GNU tar, zstd, and ~16 GB disk space for study data

Setup

Core

Install Python package dependencies to a virtual environment:

uv sync --all-groups

Activate the virtual environment:

source .venv/bin/activate

Study Data

Download rosbags.tar.zst from the Dropbox Study Data link in this file's header into data/fluentrobotics.

Extract the tar file:

cd data/fluentrobotics
tar xf rosbags.tar.zst

Repository Layout

.
├── data/
│   └── fluentrobotics
│       (Data files from our user study)
│
├── deployment_scripts/
│   (Convenience scripts used during our user study)
│
├── src/
│   ├── eleyng/
│   │   └── table-carrying-ai/
│   │       (3rd party code required for the VRNN baseline, see README.md in this folder)
│   │
│   └── fluentrobotics/
│       └── icmpc_collab_transport/
│           ├── core/
│           │   (Implementation of the core inference algorithm)
│           │
│           ├── deployment/
│           │   (ROS 2 nodes to run algorithms on a Hello Robot Stretch 2)
│           │
│           └── evaluation/
│               (Modules to analyze and visualize data from our user study)
│
├── pyproject.toml
│   (Specification file for Python package dependencies)
│
└── README.md
    (📍 You are here)

Citation

@inproceedings{yang2025implicit,
  author={Yang, Elvin and Mavrogiannis, Christoforos},
  title={Implicit Communication in Human-Robot Collaborative Transport},
  year={2025},
  booktitle = {Proceedings of the 2025 ACM/IEEE International Conference on Human-Robot Interaction},
  pages = {23–33},
}