From Metamodeling to Automatic Generation of Multimodal Interfaces for Ambient Computing

International audiencehis paper presents our approach to design multichannel and multimodal applications as part of ambient intelligence. Computers are increasingly present in our environments, whether at work (computers, photocopiers), at home (video player, hi-fi, microwave), in our cars, etc. They are more adaptable and context-sensitive (e.g., the car radio that lowers the volume when the mobile phone rings). Unfortunately, while they should provide smart services by combining their skills, they are not yet designed to communicate together. Our results, mainly based on the use of a software bus and a workflow, show that different devices (such as Wiimote, multi-touch screen, telephone, etc.) can be coordinated in order to activate real things (such as lamp, fan, robot, webcam, etc.). A smart digital home case study illustrates how using our approach to design with ease some parts of the ambient system and to redesign them during runtime