The convergence of the two technologies is a fundamental part of Industry 4.0. Through data analysis, numerous activities that improve production and enable smart manufacturing are possible
There is an almost natural correlation between the Internet of things (IoT) and big data technologies. The former are in fact producers of large amounts of raw data, the latter provide the tools to filter, sort and analyse this data and extract value from it. The convergence of IoT and big data thus constitutes a powerful tool that can be made available to companies for a variety of purposes. For example, in smart cities, where data collected by sensors distributed across the territory constitute a valuable asset from which information of all kinds, from traffic to parking usage, can be extracted. The convergence of IoT and Big data is obviously an integral part of Industry 4.0: connected objects installed in factories send data about their surroundings and their own operation, enabling a range of possibilities, from monitoring workers to controlling machines themselves.
There are different domains to which the continuous circular process generated by the convergence of IoT and big data can be applied, in a crescendo that goes from the individual production process to involving not only the manufacturing company but also its customers (in a servitisation for example). Applied to a production process, it allows it to be monitored in order to improve it and adapt it to changing external conditions. Or, the process can be adopted by different departments of the same company in order to achieve the best integration and improve the results of the entire company. If, as mentioned above, the IoT big data convergence and the continuous circular process are also extended to the customers of the company adopting these technologies, the information sent by the products will provide valuable insights into their operation and the need for maintenance.
In this respect, it should be emphasised that IoT - big data convergence enables predictivemaintenance, i.e. it allows machines and devices to be serviced before they malfunction or fail, whether they are in the factory or installed at the customer's premises.
Finally, the interconnection between objects, devices and machinery, in addition to obtaining information from them, has the advantage of being able to control or configure them remotely. A possibility that can prove extremely useful when the object to be monitored, controlled or programmed is located in an inaccessible place to reach or is placed in a dangerous or hostile environment for the worker.
Thanks to the ever-increasing miniaturisation of sensors and circuits for wireless communication, it is now possible to embed a device in almost every object to collect information about the environment and transmit it over the network.
These devices are referred to as 'embedded systems'. Added to this is the availability of multiple wireless networks to connect objects and machines. The data that IoT devices generate are different depending on the type of object or machine being considered. They can be:
The combination of IoT and AI makes it possible to develop increasingly innovative and advanced technological solutions, so much so that we can speak of 'AIoT'. But how can these two realities be integrated? Artificial intelligence exploits a very valuable resource to be able to function and constantly improve: Big Data! But for one to really be able to speak of machine intelligence, similar to human intelligence, these resources must be reliable and always available.
To overcome this problem, IoT technologies come into play. The latter are in fact capable of collecting, aggregating, analysing and providing predictive models by exploiting a very complex system of platforms and devices. Hence, the joint use of IoT and AI increases the mutual value of the two solutions: on the one hand, AI increases the potential of IoT by implementing machine and device learning; on the other hand, IoT increases the value of AI by providing resources in terms of connectivity and data exchange.
The combination of IoT and AI, which we can refer to as AIoT, thus enables even more reliable and accurate data, predictive models and functions, providing a solid basis on which to develop new efficient and effective technological solutions.
There are several ways to store data generated by IoT devices, including for analysis purposes: on premises (on-premises), in the cloud or hybridising the two options. The choice between the various possibilities depends on the volume of data, but also on the type of connectivity, as well as other factors (e.g. when there are power supply problems for the devices). Another discriminating factor is the intended purpose of the data: it makes a difference whether the data is collected for storage purposes or for real-time analysis. Alternatively, the analysis may be performed in batch.
Let us examine some methodologies of big data analytics.
