Adapting a Floor Sensor for Autonomous Robots in a Smart Building

Takashi Matsumoto, Motoaki Yamazaki, Michihito Shiraishi
We demonstrate a real time recognition of autonomous robots using a floor sensor. In these years, many types of autonomous mobility such as delivery robots and guiding robots are introduced to building facilities. To provide smooth operation of robots in a building, accurate position detection of robots by infrastructural sensors is required. By privacy concerns, there are many situations that camera recognition is not appropriate to install. Therefore, we adapted a flooring pressure sensor system of Fichvita to take positions of robots and people anonymously. This system had a function to recognize human footsteps originally. To enable robot recognition extending the function, we collected trajectory data of robots by the system installed in our building. By improving the software, now it recognizes a guiding robot and a delivery robot in addition to human footsteps in real time.

Teaching Conversational Robots in a Museum Exhibition with Interactive Surfaces

Heloisa Candello, Mauro Carlos Pichiliani, Claudio Santos Pinhanez
Artifcial intelligence concepts is incorporated in many formal and informal educational institutions. This demo showcases a novel learning experience of fundamental I concepts for children in a 30-minute, engaging, hands-on, and playful experience designed for museums and community centers. The experience aims to make children to comprehend that AI systems are built from and by people, and is materialized in a series of activities where the children improve the abilities of three AI-powered conversational robots competing in a quiz game. The demo shows how children can learn about AI by teaching machines in natural language with interactive surfaces in a way which enables them to grasp intuitively those essential ideas about AI.