Jarvis

1. Context

1) Problematic

According to the “building efficiency initiative” website [1], the main issues that smart buildings are intended to tackle consist in energy consumption, security, and maintenance. Considering the energy consumption under the broader angle of resources repartition, and being given that buildings present a wide range of resources, it is likely that an correspondingly wide range of enquiries needs to be satisfied. For big buildings such as libraries that possess plenty of work space, occupancy and comfort are critical. These very features are equally important for both library users and the managing staff, but are perceived differently. When the library user wants to have access to the room availability, the manager will have in mind occupancy and security issues, for degradation can easily occur. When the library user wants a comfortable room, the manager will have in mind the energy consumption and cost. JARVIS (Just A Real Vision of Indoor Space) is the easy-to-deploy solution we have developed to address this problem. Its purpose is to save energy by facilitating both the library user’s search and the matching of occupancy patterns with energy use for the library manager. 

Figure 1 shows a typical scenario of a user looking for a room without JARVIS, with the possibility of crossing a lot of occupied rooms before finding one available. Even then, it is not certain to have found a comfortable one (to guarantee this the manager should warm all rooms, which is neither ecologic nor economic). And once the Holy Grail has been reached, other valiant people might still seek for it and disturb the work session. Figure 2 illustrates the same quest using JARVIS, and shows that the manager’s choice of having made available comfortable rooms is more transparent for the library user who has now freedom to choose a room. Energy is saved by both teams.[pic 1]

[pic 2]

2) Overview of Jarvis

Jarvis is an easy-to-deploy solution which basically aims at giving comfort metrics to a library user as well as energy monitoring and basic security features to the management staff. To do so, several sensors send data to the cloud. Then all the data are crunched together to give a prediction of the building occupancy, which is then returned to the library user and staff (figure 3) ^[a]. [pic 3]

1. Methodology

1) Approach of the problem

It has been made clear that this problem has four key features, namely comfort, security, energy consumption, and occupancy. Comfort and energy consumption are two sides of the same coin, which edge could be the occupancy. This is good news, as this means that the correlation between these values should be good. In other words, a few sensors should be enough to get the relevant measurements, and the data analytics should work fine. The security aspect darkens the picture a little bit, as it can in itself cover a lot of possibilities. The choice we have made is simply to take photography of people entering the room and to consider that occupancy at undue hours can be perceived as a security breach. This is left to the consideration of staff members.

...