Enhancing the physical abilities of the human body is desirable for a number of reasons. These reasons include, but are not limited to, avoiding injury of workers who have to handle heavy loads in situations and environments where it is not possible to use conventional machinery (e.g. forklifts). A potential solution to this problem is the use of robotic exoskeletons to augment the strength and endurance of the human body for load-handling tasks.
This study is part of a larger, industry-funded group-research project, done with the aim of developing an enhancive exoskeleton. The needs and target requirements of the final prototype have been determined based on the market-oriented goals of the group project. An energetically autonomous exoskeleton with an acceptably high load-carrying capacity is to be developed, and the key to accomplishing this is the optimal design of the actuation system. The ideal actuation system needs to be strong, but also power-efficient, so that it can be powered by a light-weight, portable power supply. The actuators should also be lightweight, so that the total weight of the exoskeleton is low enough to be safe for the human user. Therefore, this study was done with the aim of developing an optimal design for the actuators to achieve high load-carrying capacity, and low weight and power consumption. To be more specific, the aims of this research included the identification of the degrees of freedom (DOFs) to be actuated, obtaining the torque and power requirements for each actuator, and to design the actuators using the optimal motor size and optimal power transmission mechanism.
Since initial investigations suggested the use of electric motors to achieve an untethered design, the baulk of the work done in this study is focused on actuator design using electric motors. Furthermore, the scope of this research is limited to the lower-body DOFs (namely the ankle, knee and hip joints) in the sagittal plane.
To address the above-mentioned design problem, dynamic modelling and simulation of the exoskeleton movements were performed to obtain the torque and power requirements at the joint. These requirements, in addition to being used later in a novel optimisation algorithm, were also used as guidelines for a market search on electric motors, which resulted in a list which represents the current state of the art of electric motors. The list of motors was saved as a spreadsheet, in the form of a table containing the technical data which characterise each motor. Similar tables were also created for a number of different types of power transmission mechanisms considered in this study, namely strain gears, chain-and-sprocket mechanisms, and ballscrews. These lists have been used by the optimisation algorithm, which was developed to combine the mathematical models of a motor and a transmission mechanism from the lists, assess the performance of the combination, and repeat this procedure for each and every motor and transmission mechanism in these lists. Thus, through an exhaustive search, the optimum choices for the motors and power transmissions system can be determined for each actuator. Based on the results of the developed optimisation algorithm, a single-joint test prototype was designed, built and used to perform experiments in order to test and validate the reliability of simulations used in the optimisation algorithm. The test results were also used to modify the assumed values of an efficiency parameter within the simulation program. The optimisation algorithm was then refined with the modified parameter value, and the optimal designs of the actuators were obtained for the knee, hip and ankle joints in the sagittal plane. It was also discovered that the most power-efficient motors also yielded the upper bound of the required load-carrying capacity, which is 60 kg. In addition, energy harvesting aspect of such robotic exoskeletons have also been explored.
