IoT architecture 

• ·         In IoT architecture, several building blocks are connected.
• ·         Sensor- the device is used for collecting, storing, and processed data over a network to the big data warehouse.
• ·         Actuators devices’ perform commands sent via a user application.

Fig:- IoT architecture.

Sensors:-

·         Collect data and transferred it over a network and actuators perform the action (for example, to switch on or off the light, to open or close a door, to increase or decrease engine rotation speed and more).

 Gateways:-  

·         Sensors send Data to the cloud and vice versa through the gateways.

·         It provides connectivity between things(Sensors) and the cloud and executes commands using their actuators.

 Cloud gateway:-

·         Cloud gateway facilitates data compression and secures data transmission between field gateways and cloud IoT servers.

·         It ensures compatibility with various protocols and communicates with field gateways using different protocols depending on what protocol is supported by gateways.

 

Streaming data processor:-

·         It ensures effective transition of input data to a data lake and control applications. No data can be occasionally lost or corrupted.

 Data lake:-  

·         It is used for storing the data generated by connected devices in its natural format. Big data comes in "batches" or in “streams”.

·         When the data is needed for meaningful insights it’s extracted from a data lake and loaded to a big data warehouse.

 Big data warehouse. 

·         Filtered and pre-processed data is extracted from a data lake to a big data warehouse.

·         A big data warehouse contains only cleaned, structured, and matched data (compared to a data lake which contains all sorts of data generated by sensors).

·         Data warehouse stores related information about things and sensors (for example, where sensors are installed) and the commands control applications send to things.

 Data analytics. 

·         Data analysts use data from the big data warehouse to find trends and gain actionable insights.

·         When analyzed (as visualized - diagrams, infographics) big data show (example- the performance of devices, help identify inefficiencies and work out the ways to improve an IoT system) make it more reliable, more customer-oriented.

·         Also, the correlations and patterns found manually can further contribute to creating algorithms for control applications.

 

Machine learning and the models' ML generates. 

·         Machine learning, use to create more precise and more efficient models for control applications.

·         Models are regularly updated (for example, once a week or once a month) based on the historical data accumulated in a big data warehouse.

·         When new models are tested by the applicability and efficiency and approved by data analysts, new models are used by control applications.

 

Control applications

·         Control applications send automatic commands and alerts to actuators.

·         Example: in a smart home can receive an automatic command to open or close the gate. The forecasts weather service. Soil monitoring, help monitor the state of industrial equipment, and in case of a pre-failure situation, an IoT system generates and sends automatic notifications to field engineers etc.

·         The commands sent by control apps to actuators and stored in a big data warehouse.

·         It investigates problematic cases, security breaches by control commands etc.

·         Control applications can be either rule-based or machine-learning based.

·         Control apps work according to the rules stated by specialists and use models for regularly updating (week/month) with the historical data stored in a big data warehouse.

·         Ensure better automation of an IoT system, to perform certain actions).

 

User applications

·         These are software components of an IoT system that enable the connection of users to an IoT system and give the options to monitor and control their smart things (for example, homes or cars and controlled by a central system).

·         With mobile or web app, users can monitor the state of their things, send commands to control applications, set the options of automatic behavior (automatic notifications and actions when certain data comes from sensors).

 

Device management

·         To ensure sufficient functioning of IoT devices.

·         Manage the performance of connected devices (facilitate the interaction between devices, ensure secure data transmission, and more):

·         Device identification:- to establish the identity of the device (the genuine device with trusted software transmitting reliable data).

·         Configuration and control:- configure devices according to the purposes of an IoT system (example, unique device ID). And updates.

·         Monitoring and diagnostics:- to ensure the smooth and secure performance of every device in a network and reduce the risk of breakdowns.

·         Software updates and maintenance:-  to add functionality, fix bugs, address security vulnerabilities.

 

User management

·         Provide control over the users having access to an IoT system.

·         It involves identifying users, their roles, access levels, and ownership in a system.

·         It includes such options as adding and removing users, managing user settings, controlling access of various users to certain information, as well as the permission to perform certain operations within a system, controlling and recording user activities, and more.

 

Security monitoring

·         Connected things produce huge volumes of data so need to be securely transmitted and protected from cyber-criminals.

·         To prevent it create a log and analyze the commands sent by control applications to things, monitor the actions of users, and store all these data in the cloud.

·         It monitors security breaches at the earliest stages and takes measures to reduce their influence on an IoT system (for example, blocking certain commands coming from control applications).

·         Identify the suspicious patterns of behavior, store these samples and compare them with the logs generated by IoT systems to prevent potential penetrations and minimize their impact on an IoT system.

 

In brief:-  

In simple terms, our IoT architecture contains the following components:

• Things equipped with sensors to gather data and actuators to perform commands received from the cloud.
• Gateways for data filtering, preprocessing and moving it to the cloud and vice versa, – receiving commands from the cloud.
• Cloud gateways to ensure data transition between field gateways and central IoT servers.
• Streaming data processors to distribute the data coming from sensors among relevant IoT solution components.
• Data lake for storing all the data of defined and undefined values.
• Big data warehouse for collecting valuable data.
• Control applications to send commands to actuators.
• Machine learning to generate the models which are then used by control applications.
• User applications enable users to monitor control their connected things.
• Data analytics for manual data processing.
• Device and User management components to provide stable and secure functioning of things and control user access issues. For this use some  IT solutions such as ERP, MES, WMS, delivery management systems, and more.

 

===================================================================

 

IoT architecture example – Intelligent lighting in smart home.

Basic components

·         Sensors take data from the environment (example, daylight, sounds, people’s movements). Lamps are equipped with actuators to switch the light on and off.

·         A data lake stores raw data coming from sensors.

·         A big data warehouse contains the extracted info smart home behaviour on various days of the week, energy costs, and more.

 

Manual monitoring and manual control

·         Users control smart lighting system with a mobile app .

·         With the app, users can see which lights are on and off and send commands to the control applications that further transmit them to lamp actuators. An app can also show which lamps are about to be out of order.

 

Data analytics

·         Analysing smart lighting, according their, and other info gathered with sensors, data analysts can make and update the algorithms for control applications.

·         Data analytics also helps in assessing the effectiveness of the iot system.

·         For example, if a user switches off the light right after a system automatically switches it on and vice versa.

 

Automatic control’s options and pitfalls

·         The sensors monitoring natural light send the data about the light to the cloud. When the daylight is not enough, the control apps send automatic commands to the actuators to switch on the lamps.

·         The rest of the time the lamps are switched off.

·         Extraneous light captured by sensors can make the smart system conclude that it’s enough light, and lighting should be switched off.

·         Thus, it makes sense to give the smart system a better understanding of the factors that influence lighting and accumulate these data in the cloud.

·         When sensors monitor motions and sounds, it’s not enough just to switch on the light when movements or sounds are identified in the room or switch all the lamps off in the silence. Movements and sounds can be produced, for example, by pets, and cloud applications should distinguish between human voices and movements and those of pets.

·         If the noises coming from the street and neighbouring houses and other sounds then it’s possible to store the examples of various sounds in the cloud and compare them with the sounds coming from the sensors.

 

Machine learning

·         Intelligent lighting can apply models generated by machine learning, example, to recognize the patterns of smart home owners’ behaviour (leaving home at 8 am, coming back at 7 pm) and accordingly adjust the time when lights are switched on and off (for example, switch the lamps on 5 minutes before they will be needed).

·         Analysing users’ behaviour in a long-time perspective, a smart system can develop advanced behaviour.

·         Example, when sensors don’t identify typical movements and voices of home inhabitants, a smart system can “suppose” that smart home residents are on a holiday and adjust the behaviour: for example, occasionally switch on the lights to give the impression that the house is not empty (for security reasons), but do not keep the lights on all the time to reduce energy consumption.

 

User management options

·         To ensure efficient user management, the smart lighting system can be designed for several users with role distribution: for example, owner, inhabitants, guests.

·         In this case, the user with the title “owner” will have full control over the system (including changing the patterns of smart light behaviour and monitoring the status of the yard lamps) and priorities in giving commands (when several users give contradicting commands, an owner’s command will be the one control apps execute), while other users will have access to a limited number of the system’s functions. 

·         “People” will be enabled to switch on and off the lamps with no opportunity to change settings. 

·         “Guests” will be able to switch on and off the light in some parts of the house and have no access to controlling the lights, for example, near the garage.

 

=============================================================