# SocNav1: A Dataset to Benchmark and Learn Social Navigation Conventions

arXiv: https://arxiv.org/abs/1909.02993 · HTML: https://mlsanity.com/paper/1909.02993/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Sep 2019.

## Built on by (most cited first, 10 of 10)

- [Socially Compliant Navigation Dataset (SCAND): A Large-Scale Dataset of Demonstrations for Social Navigation](https://mlsanity.com/paper/2203.15041/) (2022)
- [Principles and Guidelines for Evaluating Social Robot Navigation Algorithms](https://mlsanity.com/paper/2306.16740/) (2023)
- [Federated Continual Learning: Concepts, Challenges, and Solutions](https://mlsanity.com/paper/2502.07059/) (2025)
- [Federated Continual Learning for Socially Aware Robotics](https://mlsanity.com/paper/2201.05527/) (2022)
- [A Graph Neural Network to Model Disruption in Human-Aware Robot Navigation](https://mlsanity.com/paper/2102.08863/) (2021)
- [Learning Social Navigation from Demonstrations with Conditional Neural Processes](https://mlsanity.com/paper/2210.03582/) (2022)
- [SocNavGym: A Reinforcement Learning Gym for Social Navigation](https://mlsanity.com/paper/2304.14102/) (2023)
- [Socially aware navigation for mobile robots: a survey on deep reinforcement learning approaches](https://mlsanity.com/paper/2512.00049/) (2025)
- [Federated Learning of Socially Appropriate Agent Behaviours in Simulated Home Environments](https://mlsanity.com/paper/2403.07586/) (2024)
- [LocoVR: Multiuser Indoor Locomotion Dataset in Virtual Reality](https://mlsanity.com/paper/2410.06437/) (2024)

## Abstract (arXiv)

Adapting to social conventions is an unavoidable requirement for the acceptance of assistive and social robots. While the scientific community broadly accepts that assistive robots and social robot companions are unlikely to have widespread use in the near future, their presence in health-care and other medium-sized institutions is becoming a reality. These robots will have a beneficial impact in industry and other fields such as health care. The growing number of research contributions to social navigation is also indicative of the importance of the topic. To foster the future prevalence of these robots, they must be useful, but also socially accepted. The first step to be able to actively ask for collaboration or permission is to estimate whether the robot would make people feel uncomfortable otherwise, and that is precisely the goal of algorithms evaluating social navigation compliance. Some approaches provide analytic models, whereas others use machine learning techniques such as neural networks. This data report presents and describes SocNav1, a dataset for social navigation conventions. The aims of SocNav1 are two-fold: a) enabling comparison of the algorithms that robots use to assess the convenience of their presence in a particular position when navigating; b) providing a sufficient amount of data so that modern machine learning algorithms such as deep neural networks can be used. Because of the structured nature of the data, SocNav1 is particularly well-suited to be used to benchmark non-Euclidean machine learning algorithms such as Graph Neural Networks (see [1]). The dataset has been made available in a public repository.