Immersive Realities Hackathon 2025: XR meets AI
13–15 November 2025 | Munich & Online

The Immersive Realities Hackathon 2025 is a dynamic, interdisciplinary event where students from Bachelor to PhD level come together to explore the future of immersive technologies powered by AI. Organized by the Immersive Realities Working Group at the TUM Think Tank, the hackathon connects academia, industry, and policy through hands-on innovation. Over three days, participants form teams, receive mentorship from academic and industry experts, and prototype solutions at the intersection of extended reality, artificial intelligence, and social impact. Core topics include ethical AI in XR, gamification in AR, and the design of socially aware virtual agents. With awards for technical excellence, innovation, and ethics, this hybrid hackathon empowers students to launch startups, build careers, and contribute to responsible tech development. Whether you're passionate about digital futures or eager to shape them—this is where your ideas come to life.

The event is co-organized with XR Hub Bavaria, AI Grid, Women in AI, Stelldirvor, and the Institute for Ethics in Artificial Intelligence. 

AGENDA

Thursday, 13.11.25

• 10:00-22:00     TUM Think Tank open as student hub (all rooms) 
• 10:00-10:45     Kick Off Session // Challenges will be announced (45min, hybrid)
  Zoom-Link: Opening Event 

Join Zoom Meeting

https://tum-conf.zoom-x.de/j/7980657712?pwd=bEYwUWRWQ0NnZmUxWFhWT1laU1l6dz09    Meeting ID: 798 065 7712 Passcode: 657417

• 14:00-14:30     Walk-In Session / Check in for students (hybrid) - Link 
  
• 15:00-17:00     Workshop (onsite, closed-door)

Friday, 14.11.25

• 09:00-22:00     TUM Think Tank open as student hub (all rooms) 
• 10:00-10:30     Walk-In Session / Check in for students (hybrid) - Link 
• 15:00-17:00     Workshop (onsite, closed-door)

Saturday, 15.11.25 

• 08:00               End of Hackathon (stop submission) 
• 09:00-12:00     Jury grading the projects 
• 13:00               Closing Event & Award Ceremony

VENUE

The hackathon will take place at TUM Think Tank, 1st floor in Richard-Wagner-Straße 1, Munich. For participants joining online, the open and closing events will be streamed via Zoom. The Link you can get above!

We will have three rooms for workshops and working space for students, mentors, and partners. Rooms are open from 9:00 to 22:00.

Available onsite rooms:

• B.158 & B.160 – participant working spaces 
  B.162 – mentor work space 
• Balcony – social space  
• If you are TUM Student, you can access StudiTUM 24h.

Discord Channel: XR & AI TUM ThinkTank Hackathon

Mentor Overview

Here you will find an overview of the mentors as well as the links to their personal zoom room Mentor Overview

How to submit: How to submit your project on devpost 

Useful Resources

A link to some useful resources 

FAQ Document

Opening Event Slides:  Opening Event Slides

Challenges

AI in Immersive Environments 

Challenge 1 - Adaptive Training, Cross-Device

Design a cross-device XR training experience that uses AI to personalise human learning and includes multisensory feedback in real time, while predicting and optimising latency to keep immersion consistent.

Pick any learning context (e.g., hard skills, soft skills) and propose a user-centric pipeline: input signals (performance, errors, pace, optional biometrics), an adaptation policy, and, if possible, a lightweight prediction/optimisation loop that manages latency across VR/desktop/AR. Demonstrate how you’ll measure learning gains, and validate adaptation. Address ethics by design and sketch a business case. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief. 

Challenge 2 - Emotion-Aware, Responsible by Default

Create an emotion-aware XR environment that adapts content and agent behaviour responsibly within legal and ethical limits, demonstrating measurable user benefit and a credible path to economic value.

Choose a purpose and define an emotion inference approach that adapts ambience, pacing, or agent responses. Make safeguards explicit. Specify accuracy and harm-reduction criteria and outline beyond-compliance governance. Close with a succinct business model. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief. 

Social Agents in Immersive World

Challenge 1 - Private-by-Design Self-Representation

Build a customizable avatar system that allows users to decide how they are represented (in terms of appearance, mood, and intention cues), while cryptographically separating their real-world identity from the avatar and ensuring bias-aware generation and responses.

Design an identity pipeline where a clear name is never involved in avatar generation and where bias checks are built in. Let users set a mood/intent scale that drives visual cues and agent reactions, with transparent indicators and an always-on “manual mode.” Define evaluation: privacy threat model, fairness metrics, and usability. Show business viability with at least one target segment and a lightweight compliance note. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief. 

Challenge 2 - Ethical Multi-Agent “Village”

Create a small, XR “village” where autonomous virtual agents (two or more) interact with each other and with humans under transparent social norms, demonstrating ethical behaviours, community health, and clear user value.

Specify agent roles, goals, and norms, and demonstrate how agents’ policies adapt while staying within the red lines. Include human-in-the-loop controls, event transparency, and safeguards for vulnerable states. Define evaluation: community metrics, user-centric outcomes, and simple safety telemetry. Outline a business case with a minimal pricing/deployment sketch. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief. 

Gamification in AR

Challenge 1 - AR Safety Coach for High-Risk Work

Build an AR training + coaching experience that uses game mechanics to improve safe performance in high-risk settings (e.g., hiking, manufacturing, operating heavy machinery etc.) and proves measurable impact without compromising rights or safety.

Pick one scenario and map a pipeline from offline rehearsal to real-life assistance on phone/tablet/HMD. Use mechanics that reward correct processes over speed, and never gamify risk. Define evaluation and ethics by design. Close with a business sketch. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief. 

Challenge 2 - Everyday AR for Behavioural Transformation & Inclusion

Design an AR experience that gamifies pro-social behaviour in everyday apps to deliver a clear societal benefit and built-in inclusion - then show how you’ll measure and sustain that change.

Choose a domain and define mechanics that inform without manipulating. Bake in inclusion from day one and set evaluation targets. Cover ethics, and outline a business path. Deliver a prototype or mock-up and/or video (free of choice), your business case, evaluation metrics if relevant and an ethics brief.

• Creativity & Originality
  The idea clearly addresses the posed challenge and the XR+AI theme, showing strong novelty and inventiveness.
• Societal Impact & Interdisciplinarity
  The project demonstrates clear potential for public or user benefit and shows evidence of cross-disciplinary thinking, combining technology, design, and policy.
• Ethics & Responsible Innovation
  It follows privacy-by-design principles, ensuring inclusivity, safety, and fairness, while identifying potential risks and proposing concrete mitigations.
• Technical Quality of Prototype
  The project demonstrates a working demo or a credible prototype path, achieving the right balance between ambition and reliability while handling technical difficulty well.
• Commercial Viability & Go-to-Market Readiness
  It defines a clear customer segment with a quantified pain point, supported by a credible business model and a realistic go-to-market plan with near-term milestones and success metrics.
• Answers the Challenge Given
  The solution directly and effectively addresses the challenge provided.