WORKSHOP FRAMING
As robots increasingly enter social and domestic environments, their ability to generate context-aware and expressive behaviors becomes essential for meaningful and believable human-robot interaction. Traditional behavior generation often relies on static rules or scripted routines, limiting adaptability and nuance. In this workshop, we present a novel system that leverages Large Language Models (LLMs) to generate dynamic, embodied responses for robots. Our system integrates a defined robot persona, an affordance profile based on physical capabilities, and a response mapping framework that converts conversational inputs into sequences of expressive movement markers. Participants will engage in a hands-on experience with this system, beginning with an introduction to behavior generation in robotics and a live demonstration. Working in small groups, they will design their own prompts and response scenarios for daily life interactions, implement them using the framework, and test the results directly on a quadruped robot.

WHAT  TO EXPECT
• A comprehensive understanding of how LLMs can be leveraged to generate real-time robot behaviors.

• A guided hands-on session of learning, ideating, and implementing your own representation of robot behavior generation.
• A chance to interact with and work on the Spot robot by Boston Dynamics.

• Generate new ideas for your thesis, research work, or just another fun project!

HOW TO PARTICIPATE
Please register before October 4th by submitting the Registration Form

Please note that you must be registered to the Mindtrek conference with a 'visitor/author' pass or a 'workshop' pass.

WORKSHOP STRUCTURE
• Introduction to the Workshop (09:30 – 09:50):The day begins with an overview of the workshop’s aims, context, and structure. We introduce participants to the broader themes of expressive robotics, embodied interaction, and the role of LLMs in enabling context-aware behaviors in robots.

• Icebreaker and Getting to Know Each Other (09:50 – 10:15):To build a collaborative and open atmosphere, we engage participants in interactive icebreaker activities. These will help attendees get acquainted, understand each other’s backgrounds, and feel comfortable exchanging ideas throughout the day. 

• Coffee Break (10:15 – 10:30): A short break with refreshments.

• System Walkthrough and Demo (10:30 – 11:15): Participants are introduced to the key components of the framework: robot persona, affordance mapping, LLM prompt structure, behavioral marker system, and execution pipeline. A live demo with the Boston Dynamics Spot robot will show how the system responds to voice inputs and translates them into expressive physical behavior. Do’s and dont’s.

• Make groups (11:15 – 11:20) 

• Brainstorming (11:20 – 12:00) Brainstorm kick-off. Start thinking of scenarios and personas.

• Lunch Break (12:00–13:00)

• Hands-On Group Activity for Designing Robot Expressions (13:00–15:00): Participants are divided into small groups and guided through the process of creating their own expressive robot interactions. 

• Coffee Break (15:00–15:15): A short break with refreshments. 

• Presentation and Group Evaluations (15:15–16:45): Each group presents their scenario, LLM prompt, and resulting robot behavior. Live demonstrations are run, and peer evaluation is conducted based on expressiveness, contextual relevance, and social readability of the robot’s responses. 

• Closing Discussion and Reflection (16:45–17:00)

ORGANIZERS
Eshtiak Ahmed is a doctoral researcher at Tampere University, Finland. His current research focuses on exploring and leveraging AI to enhance and rationalize human-robot interaction (HRI) and companionship (HRC) in daily life scenarios.

Bakhtawar Khan is a doctoral researcher at Tampere University, exploring humans, technology, and more-than-human ecologies within the Finnish Flagship UNITE Programme. Her research blends AI, XR/VR, and other platforms to investigate how technology-mediated nature experiences can reshape the future and well-being.

Jiangnan Xu is a researcher in Human-AI interaction in the Gamification Group, with interests in natural language processing, multi-agent systems, and AI-assisted creativity. Her recent work focuses on leveraging large language models to support co-creative role-playing game (RPG) development, integrating narrative generation with agent-based reasoning, and explores how AI can enable collaborative storytelling and emergent gameplay in open-ended environments.

Linas Kristupas Gabrielaitis is a doctoral researcher at the Gamification Group, Tampere University, Finland. His research primarily focuses on games-playing and diagram-mapping as ficto-speculative practices for more-than-human design.

Oğuz ‘Oz’ Buruk is an Assistant Professor of Gameful Experience at Tampere University, Finland. His research focuses on designing gameful environments for various contexts such as body integrated technologies, computational fashion, posthumanism, urban spaces, extended reality and nature. He frequently employs methods such as speculative design, design fiction and participatory design.

Juho Hamari is a Professor of Gamification at the Faculty of Information Technology and Communications, Tampere University, leading the Gamification Group. Dr. Hamari’s and his research group’s (GG) research covers several forms of information technologies such as games, motivational information systems, new media, peer-to-peer economies, and virtual economies. He is among the most cited scholars in the world, publishing in a number of prestigious venues.

CONTACT
Eshtiak Ahmed: eshtiak.ahmed@tuni.fi

Bakhtawar Khan: bakhtawar.khan@tuni.fi

Jiangnan Xu: jiangnan.xu@tuni.fi

Linas Kristupas Gabrielaitis: linas.gabrielaitis@tuni.fi