To fuel AI reasoning, vast amounts of high-quality data are
required– largely biometric, generated by the internet of things (IoT) sensors, from which personal behavioural patterns can be decoded. The current data science approach resides on the concept of collecting data into centralised data hubs for analysis. It combines the technologies of smart sensors streaming data through high-performance networks, storing vast amounts of data in data lakes and using reasoning on data with supercomputers.
Data collecting concepts provide essential property rights in the hands of data collectors and
putdataproducersinthepassiverolewithonlylimitedornorights(Wagner,2020).Research,
based on data collecting raises significant ethical and social concerns (Letheren et al., 2020;
Mobilio, 2020; Nath and Sahu, 2020).
Currently, data collecting concept is formatted to fill the AI needs for detailed data, where IoT
dataarecollected andstoredincentraldatalakes,excludingdataproducersatthefirststepin
the process (Perko, 2020).
IoT biometric data collecting: Currently, data are collected in big data siloes, utilised by organisations capable of storing
and analysing big data, such as governments, large companies and research organisations,
following different goals from social control, profit-related activities and open research.
3.1 Data collecting
In data collecting concept, data consumers, such as information communication technology
(ICT) services providers, social media companies or security organisations and researchers,
govern data-related processes, mostly due to their direct interest to utilise data for achieving
their organisational goals, ranging from weather predictions, marketing, smart city
management, drug development, selling data and analysis results, etc. They are
optimising data collecting, storing, data analysis and data use with the premise to
optimise the process and raise data quality properties. Since data are considered an asset for
data consumers, the collected data are rarely shared, whereas data producers are excluded
fromtheprocess.Theprocessedinformationintheformofprediction,prescriptionsormental
models is used in internal processes or shared with other stakeholders in HyR ( Hybrid reality is a brief, dynamic unstable period in which people and artificial intelligence (AI) technology coexist and affect each other.). Moreever,data producers are
generally completelyunawareoftheoperationaldynamicsofdatacollectingandareunableto
actively participate in the HyR interactions.