Design & development of an assistive spoon for persons with pathological tremor I - sensing & filtering

• Main Author: Wijaya, Ricky
• Other Authors: Ang Wei Tech
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Mechanical engineering::Assistive technology; DRNTU::Engineering::Mechanical engineering::Robots; DRNTU::Engineering::Mechanical engineering::Bio-mechatronics; DRNTU::Engineering::Mechanical engineering::Mechatronics
• Online Access: http://hdl.handle.net/10356/64032
• Institution: Nanyang Technological University

Pathological tremor, ranging from 4 to 12 Hz in frequency, is the most common movement disorder in human that affects only certain muscle groups, including distal and proximal parts of upper limbs. More than half of patients suffering from pathological hand tremor feel depressed and frustrated as they could not perform simple daily life activities by themselves, such as eating food using cutleries. Hence, a handheld assistive spoon is developed to help people with such medical conditions eat more easily. The device is capable of sensing the motion of tremor, distinguishing the erroneous motion from noise and intended motion, as well as manipulating its tip in real-time to compensate the tremor motion. This research focuses on the development of motion sensing and signal filtering system of the assistive spoon. Pathological tremor motion is sensed by a motion sensing system comprising of three triple axis accelerometers, which are well-calibrated and placed on a certain configuration. Sensing algorithm for such accelerometer placement is developed based on differential sensing kinematics; therefore, the motion sensing system is capable of producing a complete six degrees of freedom (DOF) information. Filtering of noise and intended motion out of the tremor motion as well as real-time signal estimation and prediction are performed by a previously developed modified–bandlimited multiple Fourier linear combiner (modified–BMFLC) algorithm. Consequently, a clean and real-time tremor signal is produced as the output of motion sensing and filtering system, which is indeed the input of manipulation system.