AAIS - Advanced Adaptive Intelligent Systems
The aim is to develop adaptive social robots that can understand humans’ communicative behaviour and task-related physical actions, and adapt their interaction to suit. We aim to investigate and demonstrate fluid and seamless adaptation of intelligent systems to users’ context, needs or preferences.
This project aims to achieve seamless adaptation of intelligent systems so that they can better fit the particular context, needs, or preferences of a specific user while requiring minimal or no human operator intervention. This goal contrasts with existing user-adaptive systems that typically require continuous user input and explicit personalization. The result will be a system that can perform real-time adaptation in response to user action, behavior or requests, while proactively adjusting its performance by reflecting on the effects of past actions on users’ behavior. The main contributions of the project are:
- Understanding and Design for the Smart Kitchen: Design guidelines, through participatory and inclusive HCI methods, for adaptive intelligent systems in smart homes. Critical evaluation of the value and impact for all of the deployment of adaptive social robots in assistive home environments.
- Perception and Representation of Long Term Human Action: Develop a sensor fusion framework using new perception modalities (environmental sounds, thermal sensors, pupil dilation, etc.). Introduce a representation of human action enabling real-time adaptation of plans and verbal instructions.
- Input, Output, and Interaction for Smart Assistive Technologies: Develop adaptive speaking style and non-verbal expression to allow for efficient information transfer and social presence. Introduce representations of task execution that allow for adaptation to user preferences, enabling robots to show intention, achieve transparency and improve collaboration in shared physical tasks.
Iolanda Leite (Project leader)
Funding: KTH Techlab
Duration: 2019-01-01 - 2024-12-31