Smart Adapt

In an increasingly dynamic, intelligent and decentralized open world, the technology and the types of applications evolves towards the creation of ecosystems made up of a myriad of heterogeneous and distributed devices and services, of  clear mobile and ubiquitous nature, and in a continuous technological evolution. Thanks to the infrastructure provided by the Future Internet (FI), the Cloud Computing, and the progress of mobile devices (smartphones, tablets, etc.) and wearables (glasses, smart watches, sensors in clothing, etc.) it is feasible to develop sophisticated new applications and services with a high degree of volatility through the cooperation among interconnected devices (or objects), unknown at design time, for a society of users whose needs evolve and change quickly.

Facing this situation, it is needed to develop systems that are capable of continuously adapting its behaviour at runtime to new environmental conditions, unpredictable situations, changing user needs, new devices and technologies to interact or new services to consume. This adaptation should be managed in an autonomic way since it does not seem feasible to assume the active involvement of a developer to materialize all the adaptations at runtime. However, an increasingly interesting topic is to analyze how humans can participate in the adaptation loop to improve adaptation, to adjust it, or to resolve critical exceptions that may occur.

The main goal of this project is to provide methods and techniques for the systematic and productive development of self-adaptive software in the domain of the FI, where multiple heterogeneous devices and services must be adapted to achieve a coordinated interaction and also to provide users with globalized services that fit their needs. Such systems pose new technological challenges that cannot be addressed by traditional methods. Thus, this project aims to provide new solutions from a Model Driven Engineering (MDE) and Models at Runtime (M@RT) perspective. Specifically, we will provide solutions to address the evolution of these ecosystems in a context of uncertainty wiht respect to services, devices and context. We will consider the human roles in the adaptation loop (commonly called “human in the loop”) to design non-intrusive systems that are easily accepted by users. Finally, we will provide methodological support to develop these systems.

We expect to achieve technological results that consist in proposing methods, tools and model-driven techniques to improve the productivity and the development quality of self-adaptive systems with the ability to evolve in the field of the Future Internet. Finally, since this project falls within an engineering scope, the results of this project will be validated through the development of prototypes, software infrastructures and case studies, as well as experimentation in academic and industrial environments.