The process of inclusion (understood not only as the inclusion of special needs) passes through the elements of a psychological-relational and cognitive character, or behaviors mediated by cognition and metacognition. Therefore, the role of the teacher is crucial in designing learning environments that promote inclusion through the development of relational, cognitive and metacognitive potential of students and in seeking the means to appreciate their added value. This paper focuses on models that allow the analysis of the pedagogical affordances (especially cognitive) of the VW and proposes the use of the assumptions of Feuerstein's Mediated Learning Experience (MLE); the goal is to help the teacher/ mediator to foster in the student the development of cognitive potential and make her find strategies and behavior transferable in the natural world. All this in order to promote the growth of young people interested in caring mankind and the world.

The 2030 Agenda settles inclusion as a crucial goal. The index for inclusion underlines a set of resources to guide educational agencies through a process of inclusive development. One interesting model to achieve it is the Universal Design of Learning (UDL) framework, whose roots lie in the field of architecture and cognitive neuroscience. It provides options to enhance the executive functions also with the support of assistive technologies: studies have recently contributed to investigate how AI-innovated Educational Management Information Systems (EMIS), apps, and learning assessments can offer to the teachers the opportunities to differentiate and individualize learning, to diagnose factors of exclusion in education, and predict dropout, dyslexia, or autism disorder. After a discussion on the state of research and on the preparatory concepts, the chapter presents examples of AI-supported tools, and how they can scaffold executive functions; it wants also to urge a future-oriented vision regarding AI and to re-think the role of education in society.

As more and more unstructured project-based activities fill the learning time of students, there is a growing need for assessment models for educational robotics, tinkering, making and coding activities. On one hand, there is a poor understanding or an underestimation of the need for evaluation, and its ability to improve systems and learning outcomes, while on the other, it is difficult to identify or devise suitable assessment frameworks. Examples from the international context are discussed more for their potential to raise awareness than as definitive answers.