Design and implementation of a voice controlled pick and place robotic arm

Implementation, for the ALOHA Smart Industry use case, of the modules necessary for the integration of the speech module with the robotic arm, the addition of the perception functionalities and robot actuation to realize a demonstrator showing all these functionalities working together. The final demonsrator will consider static objects and must be able to sense and find a way to manage specific objects within the environment, accordingly with the input voice command coming from the KWS algorithm, running on the SensorTile board. The operator can define two possible states for the picked object: 1) "Yes": the object is recognized as correct and is moved to its relative dropbox; 2) "No": the object is recognized as faulty and is moved to its relative dropbox; 3) "Go": once all the objects have been managed, the operator can start a new computation saying "Go" or can stop the exectution with all the other available keywords.