Playing RoboSoccer with Machine Learning: A learning and teaching experience based on STEM

The emergence of the Internet and the popularity of mobile devices is transforming education. The new student generations pose challenges to the academic models of how to acquire and share knowledge, interact with colleagues and especially how to be evaluated. An crucial challenge for programming courses is to propose projects that create opportunities for students to develop both technical and soft skills necessary in their future professional careers. The objective of this work is to share the experience of teaching and learning through a project based on STEM education in an Artificial Intelligence course in a technical university program. An approach based on active learning was used to develop the programming project. In giving teams of students a problem to solve, the project promoted the principles of active learning. Once aware of the challenge ahead and what skills they would need, students were offered workshops to prepare them with the skills necessary to accomplish the final task. Students were given the main materials required to complete the project, namely a small programmable robot powered by a smartphone. The evaluation was based on rubrics and oriented towards a functional product. Each team developed a robot prototype that could to participate in match of RoboSoccer and made a technical report about the project. The first results showed that the students were working in an integral way, professional competences related to teamwork, research, creativity, communication and leadership, mainly. Also, the teams were able to develop a robot prototype that could play a RoboSoccer match.