The global aging population continues to grow rapidly, bringing significant challenges to healthcare, rehabilitation, and assisted living systems. Older adults often experience progressive physical, cognitive, and mental health changes that affect mobility, independence, and quality of life, increasing vulnerability to injury, disability, and chronic conditions. Supporting safe, active, and independent aging in clinical care, community settings and long-term care facilities requires scalable, continuous, and personalized approaches that go beyond traditional models of care.

Artificial intelligence, digital health and assistive technologies offer promising solutions by enabling continuous monitoring, rehabilitation, and early detection of health risks using multimodal data sources, such as wearable sensors, ambient devices, video, audio, and clinical records. However, translating these data into actionable insights remains challenging. Aging-related data are often sparse, noisy, imbalanced, and highly personalized, with rare but safety-critical events such as falls, agitation, or rapid health decline. Moreover, real-world deployment raises additional concerns related to model robustness, interpretability, privacy, fairness, and ethical use of sensitive health data, particularly for vulnerable populations.

In line with IJCAI guidelines, we welcome submissions exploring the use of machine learning, deep learning, large language models, agentic AI, and foundation models in aging and rehabilitation research, such as clinical text analysis, multimodal data analysis, decision support, and personalized care. Authors are responsible for ensuring the accuracy, originality, and transparency of their work and must clearly document the role of such models within their methodology.  

Submission Guidelines

All papers must be original and not simultaneously submitted to another journal or conference. The paper should be submitted to the ARIAL Workshop using EasyChair link in the LNCS/CCIS one-column page format. We will accept long papers (12-15 pages) and short papers (not less than 6 pages). The paper review process is double-blind. Authors must not write their names or affiliations in the paper and should not use self-identifying information in the paper.

List of Topics

ARIAL will also be hosting the MAISON-LLF Data Challenge (https://sites.google.com/view/arial2026/maison-data-challege)

• In this workshop, we invite previously unpublished and novel submissions in the format of regular papers, in the following areas (pertaining to aging and rehabilitation), but not limited to: 

• End-to-end data pipelines for aging and rehabilitation research, including multimodal data collection, annotation, curation, sharing, and harmonization.

• AI and machine learning methods for continuous, real-time, and long-term monitoring of older adults using wearable, ambient, audio, video, and clinical data.

• Learning from limited, noisy, imbalanced, and weakly labeled data in aging and rehabilitation settings.

• Anomaly, novelty, and rare-event detection for safety-critical behaviors (e.g., falls, agitation, wandering).

• Machine learning and deep learning methods for modeling physical, cognitive, and mental health trajectories (e.g., frailty, dementia, mobility, mental health).

• Trustworthy, robust, and deployable AI systems, including uncertainty estimation, reliability, and failure detection in real-world care environments.

• Explainable and interpretable AI for clinical decision support and personalized care.

• Privacy-preserving and decentralized learning approaches, including federated learning, on-device learning, and differential privacy.

• Fair, inclusive, and ethical AI for aging populations, addressing biases related to age, gender, ethnicity, and socioeconomic factors.

• Human-in-the-loop and clinician-in-the-loop AI systems for rehabilitation, monitoring, and intervention.

• Foundation models, large language models, generative AI, retrieval-augmented generation (RAG), and agentic AI for aging and rehabilitation, including personalization and adaptation to individual needs.

• Synthetic data generation, simulation, and data augmentation techniques for rare events and underrepresented populations.

• Natural Language Processing (NLP) and multimodal interaction techniques for communication, monitoring, and engagement in elderly care.

Publication

We are hoping to get the accepted papers published in Springer proceedings. More details coming soon. 

Venue

The conference will be held in Bremen, Germany.

Contact

All questions about submissions should be emailed to shehroz.khan@aum.edu.kw/luca.romeo@unimc.it