IoT Integration Design

IoT Integration Design

The integration of information technology and operating technologies is driving powerful synergies between the physical world of smart machines, industrial operations and facilities/spaces, and the digital world of Internet of Things (IoT) platforms, applications and insights.

IoT is another fast-growing technology, which can assist you to build and implement a data-gathering network to improve systems and establish a new channel for interaction.

IoT systems use sensors to provide operational insights from the data. Intelligence is added by integrating the analytics and ML into IoT devices. The IoT is effectively connecting the digital world to the physical world. It helps to communicate and integrate information systems and the fields’ data, and it is possible to use this data in real time. The IoT is mainly divided into three layers; the physical sensing layer, IoT Integration layer (also called the middle layer), and the IoT application layer.

Physical sensing layer:

Physical sensing includes sensors. For example, those for temperature, proximity, pressure, etc. as well as development boards (also called prototyping boards) used to acquire data from the environment, like Arduino, Raspberry, Edison etc. Along with mobile phones and their sensors, such Digital Notation webpages as the microphone to receive voice commands, proximity and accelerometer, other sensors can be grouped as passive or active (passive sensors don’t require external power sources, whereas active sensors require external power to sense the external environment). Analog sensors produce a continuous signal, whereas digital sensors produce a discrete signal.

IoT integration layer:

The IoT integration layer integrates the data collected from a device, sending it to a NoSQL database along with a distributed computing platform like Apache Spark and Hadoop. Data acquired from sensors can be stored in the cloud and can be used later to create dashboards. The IoT integration layer is also called the middleware layer. This layer ensures security, quality of service (QOS), and provides IoT gateways. You can implement this layer in your DXP application by integrating IoT frameworks such as Iotivity, AllJoyn, and Eclipse-Kura, etc.

IoT application layer:

The IoT application layer collates the data and applies ML capability to the information taken from the devices. Insights and predictions, along with appropriate data and web services, are exposed to the UI dashboards and analytics application. Frameworks like Apache Spark and Apache Kafka help to manage IoT data in real time; they provide data pipelines and steaming mode to gain insight from an IoT application for the organisation. In the layer, you can use TensorFlow, Apache OpenNLP, Apache Tika, and other ML and NLP libraries for sentiment analysis, image analysis, document analysis, time-services analysis, and other processing. IoT data can help recognise trends common in a machine that help us to understand the breakdown of applications.

At each layer of the design, various ML libraries can run as needed. IoT Integration has four open-source software components designed to collect the data from external environment (i.e. sensors and boards):

3. Data stores (i.e. NoSQL databases)
4. Management (i.e. integration of IoT framework and platform), which provides the communication to collect data from a group of sensors from multiple locations
5. Visualisation (UI applications)
6. Manipulation (ML) of time series IoT data in an easy and scalable manner

IoT frameworks like ARIoT, AllJoyn, Iotivity, etc. have implemented SOA, which is beneficial in IoT integration. These frameworks provide data interoperable capabilities to the API layer. Physically sensed activities generate the events, and these events are transmitted via the communication protocol such as MQTT to the service (middleware) layer. These events are pushed via the event bus to a REST API and then the update is reflected on web portals and mobiles devices