IoT Architecture - KatharinaIoT/IoT-Project-HHZ GitHub Wiki

A fundamental and simplified model of the reference architecture is the five layers concept.

The underlying Perception Layer is collecting and identifying information from different objects via sensor devices, like RFID or barcode, which depend on the application requirement. Afterward, the gathered data is then transferred to the network layer for secure data transmission to the information processing system. In the presented case the Perception Layer consists of the Amazon Echo Dot in Iteration I and the Face Recognition in Iteration II.

The Network Layer ensures the secure transmission of information from the sensor devices to the information processing system. The transmission medium can be either wired or wireless, depending on the sensor device. The beverage requests are forwarded via Wifi to the Middleware.

The collected data is stored and processed by the Middleware Layer. Regarding Kim, this is done by cloud services.

The Application Layer manages the applications based on the collected data through the middleware layer. In addition, this data is processed graphically. Looking at our case, the collected data and defined business rules are processed and visualized.

Within the Business Layer, the required IoT system with applications and services is defined based on the business model. In this case, KIM visual preparations of the user preferences are created.

The following figure illustrates the required functionalities and platform components for the IoT application "Smart Drinking 4.0". In this context, it is differentiated between core functionalities and advances functions.

The core functions consist of Business Rules, Process Engine, Device Registry and Device Management.

• Business Rules are essential to map various use cases including exception handling. In order to prevent misuse, Rules are defined that control the consumption of beverages. This prevents, for example, under-age people from consuming alcohol or diabetics from consuming sugar-containing drinks.

• The Process Engine coordinates defined workflows of the process models and communicates with the interfaces, applications and services based on the Business Rules Layer. For instance, there are different users whose different preferences must be taken into account.

• Data Service Integration uses the IoT platform in order to ensure stable and smooth interaction with the beverage vending machine.

• Device Registry ensures a conscientious registration and administration of the devices.

• Device Management links IoT devices with IoT applications.

Furthermore, there are additional functions.

Big Data Processing / Analytics is required to implement behavior analysis based on user profile and drinking preferences. However it is not needed in the first iteration.

Besides Security functions are essentials, in order to ensure data privacy and unauthorized access.

The following figure visualizes the IoT platform, which forms the fundament on which Smart Drinking Service is based.

The Identity Management includes facial recognition. The platform forms the basis of the app "KIM - Your smart drinking buddy!"