Today there is high pressure on this industry to be as competitive as possible, in terms of automating processes, optimizing cycle times, reducing unwanted downtime, and increasing quality among others. To achieve this process improvement many companies are evolving to what is called Industry 4.0 or Smart Factories. We expect this era to be marked by advances in emerging technologies in fields such as robotics, artificial intelligence, nanotechnology, quantum computing, biotechnology, the internet of things and autonomous vehicles.

If we look at many companies, all their factories have sensors that allow them to collect large volumes of data, whether it is about product quality, line availability or process variables. With an exhaustive analysis of these variables, it would be possible to predict trends and draw optimal conclusions for each of the processes, but unfortunately, in most cases, there is not enough time for the analysis, at this point is where artificial intelligence can come into play.

In addition, in an automotive manufacturing plant, it can be use AI-powered systems to create schedules and manage workflows, to enable robots to work safely alongside humans on production floors and assembly lines and to identify defects in components that go into cars and trucks. These capabilities can help manufacturers to reduce costs and downtime on production lines while delivering better-finished products to consumers.

How can we trust AI?

In the first instance, many people think that AI is like a black-box where is not possible to know the reasons that are valued to make certain decisions. So if they do not know these reasons, they cannot trust the results that these systems provide.

In general, we are reticent to adopt techniques that are not directly interpretable. For this reason, there are some techniques being developed to solve this. There is a trade-off between the performance of a model and its transparency, shifting the IA paradigm from a black-box to a glass-box. In this context, explainable AI (XAI) is now an emerging field that aims to address how black-box decisions are made in AI systems, inspecting and trying to understand the steps and models involved in decision making in order to increase human confidence.