BlooDragu: Enhancing Motor Skills with Robotic Arm Amar Bhatt, Luke Boudreau, Lydia Hays, and Felipe Petroski Such RIT Computer Engineering Senior Design Project—Fall 2016 Millions of people are affected with involuntary or limited muscle movement due to Parkinson's or other diseases that affect the central nervous system. This work proposes a solution to alleviate the effects of this disease by providing alternate means for user’s to perform tasks. A robotic arm system was built that augments an individual's ability to manipulate items and improve the overall quality of life. Abstract System Architecture Microcontroller & Motor Control The arm system is composed of a Bluetooth armband sensor system, called Myo, that is connected to a control system to read the user’s arm movements. The control system determines how the robotic arm should move based on these movements, then it sends the appropriate signals to a microcontroller which produces the electrical signals for the motors in the robotic arm. The system was designed to be low latency. The armband wraps around the user’s forearm, and produces various sensor data. The band also produces haptic feedback so the user knows when the arm has reached its physical limits. The control algorithm interprets these signals, and translates them into duty cycles that are sent over a serial interface (USB). The microcontroller (K64F) receives data from the controller system through a Universal Asynchronous Receiver and Transmitter (UART). The control system sends data over this serial interface regarding the positions for each of the six joints in the arm. The K64F generates six square wave signals at 3.3V that connect to each of the servo motors on the robotic arm which are powered by a separate 7.2V supply. Control Algorithm Complete System Motion data obtained from the Myo armband is translated into arm movements based on the integral of the acceleration. The 3-axis gyroscope was chosen as the main source for motion data. Each axis was assigned to a group of motors which were updated based on the velocity (integral of acceleration) obtained from the sensor followed by a threshold. A specific threshold was used to determine if the user was making significant movements, and helped eliminate noise in the system. This created a more stable idle state for the arm. The integral parameters can be adjusted to allow for optimization of movement sensibility. User Interface Robotic Arm Performance & Applications Performance Highlights Interact with objects within 5mm of precision Able to grab and hold 300g of weight Two modes of operation Haptic feedback Potential Applications Augment motor skills of those who lost them due to illness. Human and robot synergy for manufacturing assembly lines Cybernetic augmentation for amputees Hazardous environment work System Cost Sensor Data Description Cost Out Cost SainSmart 6-axis Control Palletizing Robot Arm Model, SKU:20-011-409 $103.98 SainSmart Rotatable Platform for Robotic Arm, SKU:20-014-312 $17.99 K64F Microcontroller $35.00 $0.00 Laptop $300.00 Myo Armband $199.99 Mechanical Springs $3.99 Total $660.95 $125.96 The Myo armband is how the user interacts with the system directly. It possesses a variety of sensors including an accelerometer, magnetometer, and multiple EMG sensors. The sensor data is transmitted over Bluetooth to the control system. The armband fits a variety of arm sizes, and is an off the shelf product with supported drivers and development kits. Acknowledgement This work was partially supported by the Department of Computer Engineering at RIT. Dr. Melton provided advice and guidance throughout the project. Dr. Ptucha graciously donated a Myo device for use in this work.