Big Data, AI, Smart Building, Smart Office… at a time when data is everywhere, it is increasingly influencing the design of work environments.
Even if it is not equipped with sensors, it is possible to extract useful information about the use of a building. First of all, there is the information from the safety and/or working time monitoring systems. This information can be collected at specific points (often access control) in the building and indicates presence rate within the premises. Under the GDPR, this personal data is not directly accessible to all and must not be kept for more than three months. This data must therefore be anonymized in order to be processed in the context of real estate projects.
A second source of accessible data is provided by IP addresses. Depending on their aggregation, it is possible to provide information on presence or the use of premises. The allocation or not of the workspace will be decisive for the second use. For example, famous academics Benjamin Traullé and Jean-Michel Dalle used data stored for legal reasons, such as connections to a Wi-Fi network, to compare work rates of different populations (founders versus developers in particular). The analysis of this data made it possible to define the times when people started work, the evolution of work intensity, breaks, etc.
In order to be fully exploitable, access control and IP address data must pass purity tests. The most common points of attention are related to counting number of users and taking into account temporary users (visitors, service providers, contractors, etc.). To avoid confusion special duplicates (badges passed twice in a row, multiple defects of employees for IP addresses etc.)must be carefully monitored. In addition, the measurements must be carried out over sufficiently long periods to be representative and avoid the risks due to certain peaks and off-peak (school holidays, etc.). In some cases, for example, groupings between two sites, specific treatments must be applied to anticipate the uses that will change (one person counted once on each site).
Finally, the data must be interpreted in context. For example, results from WIFI connections indicate a presence rate in a building for a team. But they do not tell us what employees are doing in the work environment. Another example is the analysis of data from the space booking system. While these provide information of the reserved spaces within the building, they do not provide information of their actual use. For example, in a recent project, meeting rooms were reserved 80% of the time. However, a visual occupancy study carried out on the basis of regular visits showed that meeting rooms were only used 30% of the time. This difference is explained by behaviour (regular reservations not honored, for example) but also by the existence of very concentrated peaks. In this case, these two sources of information, combined with targeted interviews, made it possible to understand behaviours and correctly size collaborative spaces according to real needs.
Data purification and contextualization operations are essential to use in projects, especially in cases where the results are counter-intuitive. In business interviews, it is not uncommon to hear managers explain that their teams are sedentary while the data shows a high degree of nomadism. In these cases, the credibility of the results will have to be demonstrated by explaining how the results were obtained and whether all scenarios were taken into account (”Did you take into account the providers? And how were the visitors counted? Yes, but it was a period of strike” etc.). Without context, analyses rarely reflect the entire situation and use of space. For the manager who questions the results, it is a question of avoiding hasty conclusions that could be taken such as “our desks are only used 60% of the time, so it is possible to reduce them by 40%?” The tools that can be used to work on data interpretation are numerous: interviews, in-situ observations, questionnaires, in-situ measurements, workshops and are to be adapted according to each case.
While building operation data enriches users’ knowledge, it is therefore essential to complete them with field surveys, including them in a context and giving them a critical reading. Working together, qualitative and quantitative data make it possible to approach the reality of working methods; and thus to carry out transformation projects.