Tag: black-box effect, machine learning, deep learning, decision-making, Explainable AI (XAI)
Researcher: Marco Taisch
Department: DIG
Artificial intelligence will have a predominant importance in the future, making decisions that will have a considerable impact in everyday life.
Project XMANAI, under the Horizon 2020 scheme, started in November 2020 with Politecnico di Milano as one of the main collaborator: the project involves universities and research centers as well as industrial partners, such as Txt E-Solutions, Whirlpool Emea, Cnh Industrial e Deep Blue. XMANAI stands for ‘Explainable Manufacturing Artificial Intelligence’: its aim is to explain how Artificial Intelligence can be of assistance for manufacturing and of service for society while respecting European values and principles.
This is important as the decision-making process by AI is not always visible or understandable: this is also known as the ‘black box effect’, where the machine learning/deep learning algorithms are not explainable once they are computed, which raises fears of biases and mistakes among manufacturers and among the general public. Explainable AI (XAI) is an emerging field that aims at solving this issue by inspecting and attempting to understand the steps and models involved in decision making by Artificial Intelligence.
The project, taking into consideration 4 real-life cases in which Artificial Intelligence has had a positive impact on manufacturing, aims at changing the way AI is adopted by switching to a ‘glass box’ AI model, that keeps humans in the loop of the decision making process and that produces value-based explanations for manufacturers.
