Human factors considerations of industrial exoskeleton user interaction

Work-related musculoskeletal disorders (WMSD)s remain a significant and increasing health problem in the industrial sector. This is mainly attributed to many tasks still being performed manually by workers, as they require humans’ flexibility, movement capabilities and skills. There has been a growing movement in industry towards the use of exoskeletons to augment and assist human motion. Designers and engineers have encountered significant design and technical challenges which negatively impact user safety and usability. Only once these have been resolved can these devices be made commercially available for large-scale implementation in workplaces. This thesis aims to investigate specific human factor design requirements related to exoskeleton interaction and system usability to target knowledge gaps which impede industrial implementation of exoskeletons. Exoskeletons are expensive. They also have certain risks if not used correctly, and their benefits are limited to use on specific types of tasks. Therefore, Study 1 aimed to investigate the potential impact that a manual lifting exoskeleton has on risk of back injury. The NIOSH lifting equation and AnyBody Modelling System™ were used to assess the effect of reduced load and hip torque assistance on WMSD risks and spinal loading respectively. The study indicated potential positive effects of a manual lifting exoskeleton by means of reduced loading, which highlights the possible benefits of an exoskeleton for industrial applications. Study 2 investigated human kinematic and kinetics during lifting and load carriage, to guide the system control and power actuation of an industrial exoskeleton, and ultimately the interaction and impact on the worker. Four joints were examined (shoulder, elbow, hip, knee), where the 99th percentile joint angle, velocity, torque and power were determined. These data can be used to form design considerations and guidance. The study prioritised hip flexion joint for actuation assistance. Thus, an exoskeleton designed to assist with these tasks would be classified as an active trunk exoskeleton. During spinal flexion, the structures posterior to the axis of motion lengthen. Study 3 investigated the change in length of the surface of the spine during lifting and lowering to provide design guidance for the back segment of an industrial exoskeleton. Th study revealed that the change in length was dependent on lifting technique, where stoop produced the greatest elongation (72mm). The results detail a range of elongation that the back segment of an exoskeleton should accommodate to permit natural movements. The study also identified that trunk exoskeleton design must consider greater elongation of the lumbar spine than the thoracic spine. Previous research and development demonstrated the challenge to achieve both technically feasible and user-centred design exoskeletons with good usability. Study 4 and 5 involved detailed ergonomic studies of both a passive arm exoskeleton design to assist with static overhead work, and an active trunk exoskeleton designed to assist with dynamic lifting and lowering. Both exoskeletons met their primary objectives by means of significantly reduced muscle activity and perceived musculoskeletal effort on the upper limbs and lower back respectively. The exoskeletons did not have any significant adverse effects on the users. To enhance user acceptance, the passive arms exoskeleton should streamline its footprint and the active trunk exoskeleton should improve its movement transparency. This thesis has contributed to ergonomics/human factors design and evaluation of industrial exoskeletons. Commercially, industrial exoskeletons are still in their infancy. There is a wide body of research still necessary to improve the user interaction and safety. The results of this research could be used to increase use of wearable sensors and robotics as work aids, and also other applications in mainstream society.