Smart buildings and the human-machine cloud

Abstract

The Human-Robot Cloud has previously been introduced as a framework for the creation of distributed, on-demand, reconfigurable human-machine cognitive systems[1]. These systems are made up of sensing, processing, and actuation components that are not limited to a specific type of application and potentially can be extended to multiple domains and may cover spatially smaller or larger areas. In this paper, we revisit the Human-Robot Cloud architecture and present its pilot deployment on the campus of NCSR Demokritos, a research institution in Greece. In particular, our concrete deployment aims to be demonstrated in three specific application scenarios; namely, Human-Aware Smart Buildings with Energy Optimization, Security and Surveillance, and Smart Tour Guide System. In this paper, we present in detail an example implementation of the Smart Buildings scenario: a real-world application with immediate benefits in energy optimization and energy savings. Environmentally sensitive issues, such as the ground-up development of energy efficient buildings or reducing the environmental impact of the existing infrastructure, has received much attention in the past. However, the traditionally offered solutions are central, non-transferable to other infrastructure, non-scalable and suffer from single points of failure. On the contrary, in this work, which is based on a specialization of the generic Human-Robot Cloud architecture, we attempt to move beyond the industrially available solutions to meet the requirements for scalable, reconfigurable and redistributable sensory, processing, and actuation units within buildings. A set of cameras, laser range finders, and other sensors, together with a number of processing and actuation elements, including face detection, expression recognition, and people trackers, are transformed to a prototypical reconfigurable distributed extended cognitive system, which can support multiple applications in the future.