Control Library modification to improve robot interaction with its environment

Abstract

The Poppy Project aims to offer a platform accessible to beginners and experts, allowing robots to be used in multiple fields, including research and teaching. During the work with Pypot, the control library in charge of controlling the Poppy robots, several limitations have been detected. One of the main problems found in the library is the low number of sensors implemented out of the box. Since both software and hardware are open source, a series of modifications were introduced on the control library, to fulfill some specific requirements and start using the Poppy humanoid robot in some research projects. One of the main requirements was the ability to use and control new sensors and actuators added to the robot, via Pypot. The library got improvements in the image processing module, communication with IMU and Arduino boards, implementation of facial recognition, use of proximity sensors in simulation, use of custom motors in simulation, simultaneous execution of instructions in multiple robots, display of animations through the LCD screen, and a module called “behavior controller” that executes one or more functionalities in order to create behaviors.