Industrialisation approaches to explainability in manufacturing
The implementation of Artificial Intelligence (AI) boosted the production efficiency, workplace safety and customer satisfaction by the automation of the processes. But the process to expand this technology at scale is challenging starting from data acquisition, processing up, until deployment and scaling of the models.
Information transparency is a key design principle in Industry 4.0 and despite the high accuracy of powerful tools such as deep learning and reinforcement learning techniques, AI is considered to be unintelligible to the human. The “black box” approach used in AI affects negatively the trust in the system, a factor that is critical in the context of decision-making. The solution to this opaqueness is the Explainable Artificial Intelligence (XAI) that aims the transparency of the models.
This open dialog will focus on the industrialization of the AI and how explainability approaches can be effectively integrated in the AI industrialization loop.
|10:15-10:20||General introduction and opening
|10:20-10:35||Session 1: XMANAI Project presentation, we will know XMANAI which is a project that will deliver a “glass box” AI models that are explainable to a “human-in-the-loop”.
|10:35-10:50||Session 2: XMANAI MVP
|10:50-11:15||Session 3: Explainability Directions in Manufacturing
|11:15-11:25||Open Discussion with the experts from the advisory board
• Chris Decubber, The European Factories of the Future Research Association (EFFRA), Belgium;
• Michela Magas, Innovation advisor to the European Commission and the G7 leaders, Switzerland;
• Angelo Marguglio, ENG, FIWARE foundation, Italy;
• John Soldatos, STAR Technical Manager, INTRASOFT International; Greece
|11:25-11:30||Closure of the event