An open-architecture bioloid humanoid robot controller in support of developmental disability (DD) rehabilitation / Nur ‘Aqilah Zainuddin

Robot-assisted therapy has facilitated developmental disability (DD) children in performing rehabilitation modules. However, most robot-assisted therapy controllers come with closed systems which hardly support the integration of third-party components. Therefore, there is a need for more open scheme for controller design. This thesis presented a proposed controller which is aimed to extend the features of the inexpensive commercial BIOLOID humanoid robot and is compared to the manufacturer’s closed system controller named CM-530. The proposed controller was developed based on Raspberry Pi 3 model B with hardware attached on top (HAT). Given that the proposed controller is more open and provide greater flexibility to developers, a comparison in terms of average power consumption and timestamp performance of BIOLOID robot-assisted therapy using both controllers respectively are also made towards its ability to provide better support in terms of control perspectives for DD rehabilitation modules. The average power consumption for the proposed controller was 15.12% higher than CM-530 during idle mode. In dynamic mode, the proposed controller consumed 13.79% and 19.40% higher than CM-530 controller when tested for single and multiple servomotors. These is happened due to the controller’s architecture which require a Linux operating system in order to operate while CM-530 controller does not run any operating system. All results were collected using Arduino IDE and graphed using MATLAB R2015a software. The timestamp performance for the proposed controller allowed BIOLOID robot-assisted therapy to perform all five required DD rehabilitation modules while CM-530 only performed two rehabilitation modules. In conclusion, the custom-designed controller provides better capabilities regarding to its open-architecture which gives an opportunity for other developers to delve further into the structure knowledge that helps in reducing its maintenance cost. The open characteristics of proposed controller consequently will enhance the controller potential/capabilities and can be utilized by other robot-assisted therapy in future.