Data analysis

The platform does not support internal processing of data, but this is done more efficiently by components connected to the platform to e.g. automatically, detect if a function (data) has not reported as expected.

For more advanced data analysis and e.g. AI analysis takes place via external components such as works with clustering of data and machine learning to automatically e.g. detect deviations.

Data clusters processed by machine learning partners via Fiware NGSI-9.

With IoT Opens powerful frameworks for managing edge apps, simple functions can be created to analyze values ​​e.g. carbon dioxide or particle concentration and create measures such as alarms or automatic control.

Architecture
The API on the right of the figure is used to retrieve data from the platform for processing or analysis outside the platform. Both real-time and historical data are available. The API on the left can also be used for processing ”raw” data from the devices in real time.

This type of analysis and control typically takes place at the gateway (Edge computing). The platform's edge software is installed on a number of different manufacturers' equipment, such as different fiber modems, routers for indoor use or more rugged devices for urban environments or mobile applications.

If Gateway lacks support for Edge cumputing, business logic can also be executed in the cloud. IoT Open and possibly central rule execution can be hosted in the customer's network or centrally as a service. TH1NG can deliver central hosting in a data center within Sweden's borders.

Via 24/7 central operation monitoring, automatic operation alarms can be created and sent to NOC (Network Operations Center) or to resources within the customer's organization.