Whether it's a surveillance camera, an intercom or your new hyper-modern alarm clock that knows when to wake you up, connected objects are everywhere. Connected 24/7 to the Internet, they are there to make our lives easier and we often forget they are there. However, in the shadows, a danger lurks. Most of these objects are not yet designed to withstand attacks. According to a study published in early January 2016 by Arxan and IBM, 80% of IoT applications are not tested to withstand a computer attack . Two objects that communicate necessarily exchange information, which must be analyzed and processed in order to present only the relevant data to the end user. These objects, unlike ordinary objects, are loaded with sensors that regularly collect data. For example, if you take the case of a Fitbit bracelet, it can include a heart rate sensor that collects and records the wearer's data.
However, to create the " communicating " aspect, sensors alone are not enough. In addition to these sensors that collect data, you will need infrastructures that record them. These blocks can be ecuador mobile database composed of a server hosting a Web application. This is similar to desktop software, but which is run via a web page. The exploitation of this data is however subject to very precise rules of security and protection of user privacy. This data cannot be used for unsupervised commercial purposes . We cannot imagine finding in databases the times of presence and absence in an apartment, or even the conversations of the people present. Companies that have risked a policy of exploitation of data that is not very clear have been quickly opposed by consumer associations. Like recently the American electronics brand Bose which was accused of spying on its customers and which had to explain itself and quickly review its applications. The explosion in the number of connected objects must not be accompanied by a weakening of security and protection of the end user . A clear policy adapted to the product must be established from the design stage of these objects.
Connected objects, what are the issues?
But the problem comes from the very structure of the connected objects market. While large companies are present, start-ups also contribute greatly to innovation in this field. While waiting for clear regulations, we must therefore fear massive hacking attempts. The potential is too interesting for hackers to ignore it.
Machine learning and its possibilities
With machine learning, data analysis servers are becoming increasingly intelligent and can make decisions autonomously. In the industrial world, connected objects are used, in particular, to collect data from machines (vibration, temperature) in order to anticipate breakdowns or failures. Thanks to connected objects and machine learning , we can, for example, detect a breakdown on a machine in advance: this is predictive maintenance . Many applications can thus exist for these new connected objects. Machine learning suggests a renewal of our lifestyle habits and working methods. These innovations then no longer appear as simple gadgets but as real tools.martechefficient. Connected objects intend to revolutionize our daily lives. Similar to artificial intelligence (AI), they will learn from our habits and offer us ever more personalized comfort (see the video below).