Planning for Coordination of Devices in Energy-Smart Envronments
Ilche Georgievski · University of Groningen research database (University of Groningen / Centre for Information Technology) · 2013
As a sustainability property, energy efficiency is of an extreme importance, especially in environments that are heavy energy consumers, such as homes and buildings. Nowadays, homes and buildings are equipped with many devices that could be exploited in order to make them smart and energy-efficient. Our vision is to bring convergence of smart environments, energy efficiency and automated planning by proposing a planning framework for energy-efficient coordination of devices in smart environments. We establish a proof of concept confirming that automated monitoring and control of devices can lead to significant savings not only on energy, but also on the amount paid for that energy. We envision use of Hierarchical Task Network (HTN) planning due to several reasons identified in our profound overview of the most popular HTN planners. We strive to answer several research questions relating to the general design of the planning framework, the use and improvement of HTN planning, the support for users to interact with the planning framework, and the evaluation of the framework in a living-lab set up at the University of Groningen. Modern living environments, such as homes and buildings tend to be equipped with a variety of devices usually called ‘smart’ devices. The group of devices includes different sensors and actuators, where, both, the sensors and actuators, provide information about the environment, and only the actuators enable controlling the environment. Environments that embed such smart devices are called smart environments. However, embedding smart devices into physical environments surely does not mean having a smart environment by default. Without proper processing and computation of raw sensed data, the environment will not be able to smartly react to the contextual changes and occupant needs. Furthermore, consider the following important problem of sustainability. Buildings account around 40% of energy consumption in European Union and up to 50% in the United Kingdom and Switzerland, being the largest CO2 producers (EU 2010). Moreover, the energy consumption of typical industrial and commercial buildings adds up to around 30% of the total operational costs. Thus, addressing the problem Copyright c 2013, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. and making smarter use of energy in buildings will fundamentally contribute to energy and cost savings. As sensors and actuators provide only naive control of the environment, current buildings are not optimised with respect to energy consumption. The obvious research gap can be filled by new paradigms, approaches and frameworks that will create intelligent adaptations of the environment while increasing occupant comfort and keeping the environment in the most energy and cost efficient state. The necessity of information processing and computation in terms of searching, reasoning and learning in order to create sophisticated adaptations of the environment brings us in the area of Artificial Intelligence (AI). Various approaches have been explored, such as learning, affective computing, temporal reasoning, fuzzy logic, agentbased systems, event-condition-action rules, and automated planning (Sadri 2011). We find a strong motivation to develop a framework that enables smart coordination and utilisation of a variety of devices. The framework should guarantee many desired properties, such as performance, fault-tolerance and scalability in its objective to achieve energy savings. Our vision is to introduce novel methods that compose adaptations, orchestrate functionalities and accomplish goals of a building and one or more occupants.