Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Botson: An Accessible and Low-Cost Platform for Social Robotics Research

Samuel Bellaire, Abdalmalek Abu-raddaha, Natalie Kim, Nathan Morhan, William Elliott, Samir Rawashdeh

TL;DR

Botson addresses the trust deficit in disembodied AI by integrating a large language model with a physically embodied robot to generate verbal responses and sentiment-driven gestures. The system uses a low-cost hardware stack (Raspberry Pi, Arduino, 3D-printed chassis) and a lightweight prompting framework to produce concise, sentiment-tagged responses via GPT-4o, enabling real-time, affective behavior. A pilot study suggests embodied interaction increases engagement and nonverbal communication despite voice-only agents often being perceived as more helpful, highlighting the trade-offs between usefulness and social presence. Overall, Botson offers an accessible, open-source platform for social robotics research and gesture/emotion generation, with planned upgrades to LLMs, real-time APIs, and visual context integration to broaden applicability.

Abstract

Trust remains a critical barrier to the effective integration of Artificial Intelligence (AI) into human-centric domains. Disembodied agents, such as voice assistants, often fail to establish trust due to their inability to convey non-verbal social cues. This paper introduces the architecture of Botson: an anthropomorphic social robot powered by a large language model (LLM). Botson was created as a low-cost and accessible platform for social robotics research.

Botson: An Accessible and Low-Cost Platform for Social Robotics Research

TL;DR

Botson addresses the trust deficit in disembodied AI by integrating a large language model with a physically embodied robot to generate verbal responses and sentiment-driven gestures. The system uses a low-cost hardware stack (Raspberry Pi, Arduino, 3D-printed chassis) and a lightweight prompting framework to produce concise, sentiment-tagged responses via GPT-4o, enabling real-time, affective behavior. A pilot study suggests embodied interaction increases engagement and nonverbal communication despite voice-only agents often being perceived as more helpful, highlighting the trade-offs between usefulness and social presence. Overall, Botson offers an accessible, open-source platform for social robotics research and gesture/emotion generation, with planned upgrades to LLMs, real-time APIs, and visual context integration to broaden applicability.

Abstract

Trust remains a critical barrier to the effective integration of Artificial Intelligence (AI) into human-centric domains. Disembodied agents, such as voice assistants, often fail to establish trust due to their inability to convey non-verbal social cues. This paper introduces the architecture of Botson: an anthropomorphic social robot powered by a large language model (LLM). Botson was created as a low-cost and accessible platform for social robotics research.
Paper Structure (14 sections, 7 figures)

This paper contains 14 sections, 7 figures.

Table of Contents

  1. Introduction
  2. Background & Related Works
  3. Effects of Embodiment on Trust in HRI
  4. Enabling Technologies for Affective Social Robots
  5. System Design
  6. LLM Pipeline
  7. Prompting Framework
  8. Gesture Synthesis
  9. Pilot Study Results & Analysis
  10. AI Agent Helpfulness
  11. Embodied Robot Engagement
  12. Conclusion & Future Work
  13. LLM-Based Gesture Synthesis
  14. Dynamic Expression Generation

Figures (7)

  • Figure 1: Botson in Idle State (left) and Performing a Waving Gesture (right)
  • Figure 2: High-Level Block Diagram of Botson
  • Figure 3: Flowchart of LLM Pipeline
  • Figure 4: Evaluation of Helpfulness of AI Agents by Users
  • Figure 5: Evaluation of Trust in Voice-Only vs. Embodied AI
  • ...and 2 more figures