Multimodal Interactions and Explainable AI for Reflective Physical and Virtual Learning

Background and Objective

While TEL research has made strong progress in supporting cognitive learning processes, many current systems still privilege text-based and screen-based interaction, offering limited support for learning that is distributed across physical and virtual spaces. This is especially limiting for embodied and multimodal learning activities, where reflection and feedback depend on integrating data from attention, movement, dialogue, and task performance (Casas-Ortiz et al., 2023; Casas-Ortiz et al., 2026; Santos, 2016). At the same time, generative AI and LLM-based support are rapidly entering education, yet many implementations remain disembodied and only weakly connected to learner agency, metacognition, and psychomotor aspects of learning. 

This workshop brings together researchers and practitioners from international projects funded by JSPS in Japan, European Union and Indian funding agencies to further explore how multimodal, data-informed, and AI-enhanced cyber-physical learning environments can support reflective learning. Building on the success of the 1st workshop at AIED 2025 (Majumdar et al. 2025a), the ECTEL2026 miXai^learn workshop aims to synthesize current work, share infrastructures and design cases, examine implications for metacognition and learner agency, and build a future TEL research agenda for meaningful, human-centered AI support across physical and digital learning spaces. The interdisciplinary discussion would focus on the core scientific research question: How can a multimodal interaction space, supported by data-informed technologies, facilitate embodied and reflective learning activities?

Content and Themes

The call for participation (Participation form) welcomes deliberations on individual researchers' ongoing work aligned to the following topics:

1. Data collection: Papers exploring the types of data that can be collected during both physical and virtual reflective learning sessions and the methods for collecting this data.

2. Computational modeling: Design, development, and/or training of models of human physical motion, behavior, knowledge, and cognitive aspects during reflective learning sessions

3. Inferences: Studies detecting reflection skills from multimodal log data during physical and virtual reflective learning sessions.

4. Assessment approaches and ontologies: Research on how to assess students’ reflection skills with data, as well as discussions on setting criteria for the assessment of value and fairness in evaluation.

5. Intelligent personalized systems: Papers examining the development and use of intelligent systems to support reflective learning in a personalized way and the challenges faced in implementing these systems with user modeling techniques.

6. Cultural considerations: Studies on the role of culture or individual characteristics in the development, testing, and implementation of technologies to support reflective learning across physical and virtual spaces. 

7. Translation to practice: Work/studies on the barriers and challenges faced in translating research to practice, and the strategies to address these challenges.

8. Ethics, regulation, and privacy: Analyses of the ethical considerations for collecting student data during physical and virtual learning activities, as well as the steps to safeguard student privacy and ensure that teachers are aware of the appropriate ethical behavior. Discussions on how to ensure that these technologies are beneficial to students, contributions suggesting mechanisms to mitigate or exploit the observer effect, among other issues.

Participation form https://forms.gle/HrUCDvdeQBqAHQKj7

Proposed Pre-workshop Timeline 

Submission of Participation form: July 28
Participants should complete and submit the workshop participation form by July 28, 2026. They can choose to be presenters or participants in the workshop.  

Notifications of acceptance of participation: August 18
Notification of acceptance will be sent out by August 18, 2026. Accepted presenters will be invited to give an oral presentation at the workshop. We will ask presenters to agree to the recording, archiving, and dissemination of their presentations and question-and-answer sessions.

Sharing themes for case-study discussion: September 1
The program committee will peer-review and curate the participation form content. This will ensure a well-balanced and focused workshop. We will collate the submitted information in an open-access workshop website as a first step and, along with the interested participants to co-author a journal article, synthesizing the different perspectives that emerged during the workshop.

Format and Activities

We propose a half-day workshop with an expert panel, presentations, a hands-on platform session, and group activities in the afternoon on 15 September 2026. Online participants will also be supported if required. 

Expert panel

The expert panel, in collaboration with the organizing committee, will initiate the research agenda and themes emerging from the participant submissions, aligned with their own research foci. Each panelist will briefly highlight the key challenges in that area and a possible thought pointer from the perspective of TEL.  

Presentations

The format will be lightning talks, with each presenter allotted 5 minutes. The presentations will be organized into groups based on similar themes, and after each presentation, there will be a 30-minute Q&A and discussion period. The organizers will provide guiding questions to facilitate the discussions. 

Hands-on platform interactions 

Participants will engage with the LA-ReflecT platform and related cyber-physical or AI-supported learning scenarios. This activity will foreground explainability, reflection, and the integration of physical and online learning traces. 

Group activity

Participants will work in small groups to examine a design challenge or case related to reflective learning across physical and online spaces. Groups will synthesise opportunities, tensions, and design principles, and report back in a plenary discussion. This activity is intended to generate a shared research and practice agenda for the ECTEL community.  

Time: 

15 September, 2026, from 13:30 - 17:00 CEST. 

Venue: 

Centro Asociado UNED Valencia, C/ Casa de la Misericordia 34, 46014 Valencia, Spain <map>

Workshop Organizers

Rwitajit MAJUMDAR, Kumamoto University, Japan

Huiyong LI, Kyushu University, Japan

Brendan FLANAGAN, Ritsumeikan University, Japan 

Shin-Ichiro KUBOTA, Kumamoto University, Japan 

Soumyadeep BHATTACHARYA, Microsoft Research

Aditi KOTHIYAL, IIT Gandhinagar, India

Prajakt PANDE, Aarhus University, Denmark 

Olga C. SANTOS, UNED, Spain 

Irene-Angelica CHOUNTA, University of Duisburg-Essen, Germany

Acknowledgments 

This workshop is partially supported by JSPS Grant-in-Aid for Joint International Research B (24KK0051).

Reference

1. Abou-Khalil, V., Vargo, A., Majumdar, R., Magno, M., & Kapur, M. (2023). Automated assessment and guidance of project work. In International Conference on Artificial Intelligence in Education (pp. 69-73). Cham: Springer Nature Switzerland.

2. Casas-Ortiz, A., Echeverria, J., Santos, O.C. (2023) Intelligent Systems for Psychomotor Learning: a Systematic Review and two Cases of Study. Chapter 18. Handbook of Artificial Intelligence in Education, Eds: du Boulay, B., Mitrovic, A. and Yacef, J. Edward Edgar Publishing. https://doi.org/10.4337/9781800375413.00030.

3. Casas-Ortiz, A., Portaz, M., Echeverria, J., Santos, O.C. (2026) Intelligent Systems for Psy-chomotor Learning: An Updated Systematic Review and the Role of Generative AI. Hand-book of Artificial Intelligence in Education, Eds: du Boulay, B., Mitrovic, A. and Yacef, J. Edward Edgar Publishing, in preparation.

4. Majumdar, R., Prasad, P., Kadam, K., Gatare, K., & Warriem, J. M. (2023). LA-ReflecT: A Platform Facilitating Micro-learning and Its Multimodal Learning Analytics. In European Conference on Technology Enhanced Learning (pp. 731-735). Cham: Springer Nature Switzerland.

5. Majumdar, R., Gatare, K., & Warriem, J. K. (2023b). LA-ReflecT: Multimodal Learning Analytics of Micro-learning. T4E 2023 workshop.

6. Majumdar, R., Li, H., Flanagan, B., Kubota, S. I., Bhattacharya, S., Kothiyal, A., ... & Chounta, I. A. (2025a). Multimodal interactions and explainable AI for reflective physical and online learning: MiXai^ learn workshop. In International Conference on Artificial Intelligence in Education (pp. 315-320). Cham: Springer Nature Switzerland.

7. Majumdar, R., Li, H., Yang, Y., & Ogata, H. (2024). GOAL - A data-rich environment to foster self-direction skills across learning and physical contexts. Educational Technology & Society, 27(3), 61-82. https://doi.org/10.30191/ETS.202407_27(3).RP04

8. Majumdar, R., Prasad, P., & Sane, A. (2024a). Mining Epistemic Actions of Programming Problem Solving with Chat-GPT. In Educational Data Mining Conference 2024.

9. Majumdar, R., Singh, D., Chen, M.R.A. (2024b) Designing Structured Reflections for Guiding Learners' Interactions with Generative AI. In CELDA 2024.

10. Majumdar, R., Singh D., Rajendran R., Narayanan S. & Gatare, K. (2024c). LA-ReflecT platform's affordances for distributed multimodal learning analytics. In CollabTech 2024.

11. Majumdar, R., Ocheja, P., Li, H., Warriem, J.M. (2024d) LA-ReflecT Dashboard: Scaffolding Reflection on Learning with Teacher Prompts & LLM Interactions. T4E 2024 Tools track.

12. Majumdar, R., Casas-Ortiz, A., Portaz, M. Santos, O.C. (2025b) Bringing Psychomotor Learning to Online Learning Platforms: The Kenpo Learning Simulator in LA-ReflecT. Interactive event at the 26th International Conference on Artificial Intelligence in Education  (AIED 2025)

13. Mishra, S., Majumdar, R., Kothiyal, A., Pande, P., & Warriem, J. M. (2021). Tracing embodied narratives of critical thinking. In International Conference on Artificial Intelligence in Education (pp. 267-272). Cham: Springer International Publishing.

14. Portaz, M., Garcia, P., Majumdar, R., & Santos, O. C. (2024). Mastering Mind and Movement. ACM UMAP 2024 Tutorial on Modeling Intelligent Psychomotor Systems (M3@ ACM UMAP 2024). In Adjunct Proceedings of the 32nd ACM Conference on User Modeling, Adaptation and Personalization (pp. 9-12).

15. Santos, O.C. (2016). Training the Body: The Potential of AIED to Support Personalized Motor Skills Learning. International Journal of Artificial Intelligence in Education 26, 2 (2016), 730–755. https://doi.org/10.1007/s40593-016-0103-2

16. Santos, O. C., Portaz, M., Casas-Ortiz, A., Echeverria, J., & Perez-Villegas, L. F. (2023). Designing, Building and Evaluating Intelligent Psychomotor AIED Systems (IPAIEDS@ AIED2023). In International Conference on Artificial Intelligence in Education (pp. 91-96). Cham: Springer Nature Switzerland.