Jauwairia Nasis is passionate about building technologies closer to human needs. That is why when the opportunity of carrying out her doctoral research at Computer Human Interaction for Learning and Instruction Lab (CHILI) at EPFL presented itself, she took it. She is also a Marie Curie Fellow at the EU H2020 Innovative Training Network ANIMATAS. Her research in particular revolves around leveraging AI and multi-modal data to build smarter social robots in educational settings. She also get to volunteer some of my time as the Education Lead in Switzerland for the global non-profit Women in AI.
What makes human-robot interaction in educational settings different from other forms of HRI (such as in healthcare)? In an educational setting, unlike in some other areas, interaction is NOT the goal; rather interaction is only a means to a hidden goal: Learning. Since learning is a latent process, it is challenging to build social robots with the expectation to infer “if the students are engaged in the learning process or not?”
In her research, she critically assess how ‘engagement’ is currently modeled in educational HRI in the quest to build aforementioned robots that have the goal of ‘improving learning at their core’. Given that even human experts can not always know what’s best for learning in a particular scenario, this is quite a challenge for which we are trying to leverage multi-modal learning analytics including audio, video and log data; machine learning and educational theories.
In this talk, she will take you through some of the challenges in this regard, an overview of our proposed ‘productive engagement’ framework for such goal-centric social robots as well as our key findings that can provide insights to build more intelligent agents in the educational space.