Voice Controlled Spoon Feeding Robotic Arm for the Elderly

The implementation of a 6 DOF (Degrees of Freedom) robotic arm tailored for autonomous spoon feeding tasks. The robotic arm is equipped with real time mouth state detection capabilities, enabling it to assess the user’s readiness to receive food by identifying open or closed mouth states. Using Google’s pretrained Mediapipe model, the system achieves accurate mouth state identification, facilitating an adaptive feeding process. Mediapipe’s robustness allows effective mouth state detection under various lighting conditions, minimizing the impact of shadows and reflections. Additionally, the system includes Natural Language Processing (NLP) capabilities. Voice based capabilities allow users to control the robotic arm using spoken commands. voice feedback generated using NLP ensures effective communication with the user, such as notifying them about the system’s status or providing guidance during feeding tasks. If the user’s face is not detected, the system announces, “Face not detected”. Using a robotic arm, we performed tasks to transfer banana chips and moong dal from a bowl to the user’s position. It was found that the robotic arm system successfully transferred over 75% of the banana chips and 80.2% of the moong dal to the user’s mouth.