Chin , Renee Ka Yin (2007) Neural network applications in the development of artificial intelligence modules for a patient lifting robot. Masters thesis, University Malaysia Sabah.
One of the challenges that hospital nurses face is to stay fit and healthy with the heavy workload they have to endure on their day-to-day life. The workload includes the task of patient lifting and transferring. Most nurses suffer from lower back pain due to excessive load exerted on to their lumbar spine during patient lifting and transferring. A patient lifting robot is proposed and conceptually designed in this research to solve this problem. The patient lifting robot is a semi-automatic, wheeled, bed-type robot which is modeled according to the "three-men patient lifting technique". The robot arm consists of three segments. This arm is extendable or retractable and is responsible for taking and lifting the patient from bed. Each of these segments has a conveyor belting system. The robot goes near the bed and identifies the location of patient on the bed, inserts its arm between the bed and the patient, takes the patient through the sets of conveyor belts and lifts the patient from the bed. The robot is equipped with a set of sensors which forms an integral part of the control system. In addition, safety to the patient has to be ensured while transferring the patient to the robot arm and then lifting. Hence, these complex control abilities of robot are possible only by using artificial intelligence techniques. The entire control and intelligence system of the robot consists of five (5) modules, namely (i) the patient position tracking module, (ii) the automatic procedure sequencing module, (iii) the fail safe and recovery module, (iv) the danger monitoring module, and (v) the motor speed and trajectory module. The motor speed and trajectory module is developed in a separate research and is considered as a black box in the context of this research work. This research discusses the importance of each of these modules and how they are interrelated. The modules are developed using neural networks to perform their specific functions. A carefully designed data base corresponding to each of the modules is created. These data are used for training the various neural networks. The entire function of the patient lifting robot is realized by these networks. The robotic functions are constantly monitored by nurses. Provisions are incorporated for the nurses to change any sequence of robot operations when required or when malfunctions are expected. All the neural networks of the robot's intelligence system are trained for their required tasks and tested for their successful functioning using the designed data base.
|Item Type:||Thesis (Masters)|
|Uncontrolled Keywords:||patient lifting, robot, artificial intelligence technique, data base, neural network application|
|Subjects:||T Technology > TK Electrical engineering. Electronics Nuclear engineering|
|Divisions:||SCHOOL > School of Engineering and Information Technology|
|Deposited By:||IR Admin|
|Deposited On:||04 Jun 2014 10:25|
|Last Modified:||04 Jun 2014 10:25|
Repository Staff Only: item control page