The digital revolution has introduced new terms into the industrial dictionary that are often misused, and it can be difficult to identify the correct definitions. This can even lead to overlap in the minds of those hearing about certain topics for the first time.Two hot topics in recent times that everyone wants to talk about are Machine Learning e Internet of Things: In this article, we will try to define them in the correct way to provide clarity.
Machine Learning
The Machine Learning is a branch of Artificial Intelligence., consisting of various algorithms that allow to train a computer through a set of examples to be able to Make predictions about the observed phenomenon.
The applications into which it can decline are limitless: image recognition, text translation, time series forecasting, object or individual classification.
In industry, Machine Learning is applied to produce systems that can, by automating analysis of large masses of process data , provide decision support. An application example is the generation of warnings about the imminent occurrence of a failure or the monitoring of parameters to anticipate departure from the optimal operating range.
Internet of Things
The Internet of Things properly is the use of the communication protocols common to the Internet, which anyone uses to surf the Web, to connect devices.
The topic has become particularly hot first by entering many homes through home automation, then with the advent of the Cloud. The communicating objects are also smart in that they are capable of processing information to be sent or processing information provided to them from outside.
When these concepts are applied to an industrial context, we speak more specifically of the Industrial Internet of Things.
What do Machine learning and the Internet of Things have in common?
The title of the article is deliberately provocative: these two terms refer to totally different concepts.
However, they are often used together when talking about Industry 4.0 e digital transformation, because the ease of data collection provided by IoT devices, combined with the immense storage and computing capabilities affordability made available by the Cloud, paves the way for the possibility of processing by algorithms of Machine Learning.
The devices smart also pave the way for solutions on edge: data collected by the sensor is integrated with a machine learning model directly into its microprocessor. These concepts are the basis of the smart farm, where the dispersiveness of the environment makes it nontrivial to make decisions and take timely action.
The next article will be devoted to Machine Learning and Internet of Things applications that are beginning to invade the industrial market and what problems can be addressed by combining these two technologies.
