Enhancing Occupational Back-Support Exoskeletons Versatility

In the 1970s, the scientific community began addressing the relationship between musculoskeletal disorders (MSDs) and work ergonomics. Since then, many studies have been published regarding this topic. Yet, 50 years later, MSDs are still cited as the most common work-related health problem, in the majority of the cases due to back pain. Workers performing manual material handling (MMH) activities (e.g., package loading and unloading in a warehouse or luggage handling in airports) are among the most exposed to risks and injuries. Several solutions have been proposed but, due to high implementation costs or lack of acceptability by the workers, MSD problem has not been solved. Occupational back-support exoskeletons have been investigated for their ability to reduce muscle activation and, hence, mitigate the risk of injuries. Exoskeletons can be categorized as soft or rigid, if how the assistance is delivered to the user is taken into account, or as active, passive or quasi- passive, if, instead, focus is on the actuators that generate the assistance. Regardless of the categories, exoskeletons have been designed and tested, mostly, for static bending or lifting tasks. This is justified by the high risk of developing injuries associated with these tasks, but, on the other hand, does not address the complexity that out-of-the-lab environments might present. In this regard, an interesting study reported that if passive exoskeletons are used in non-lifting tasks, as walking, rather than helping they actually hinder and obstacle the wearers’ movements. And, it is clear, that in industrial settings, workers do not lift for the whole duration of their schedule. Other relevant activities might be pushing, pulling or carrying. Indeed, the International Standard ISO 11228 establishes ergonomic guidelines also for these activities, highlighting that they could yield risk of injuries. Therefore, designing exoskeletons that address also these activities might help promote the acceptability of these devices. In this work, exoskeleton versatility is defined as the device ability to recognize which activity the user is performing and provide assistance accordingly. As an example, assisting with lifting could require the design of strategies that are not fit for carrying assistance. More in details, in the first part of this thesis, it is shown that considering spinal loading, carrying has an impact comparable to that of lifting. However, if lifting control strategies are used to assist with carrying, this produces gait impairments. The need of assisting with these two activities in different ways requires, first of all, a mean to recognize which task is taking place. This is addressed in the second part of the thesis. In particular, by using Support Vector Machines (SVM), it is shown that it is possible to build an automatic Human Activity Recognition (HAR) algorithm. Moreover, this algorithm can be embedded in the control architecture of XoTrunk, the back-support exoskeleton developed at our lab, enhancing its versatility. The third and last part of this this work assesses the advantages of using a versatile exoskeleton, comparing its performance to state-of-the-art exoskeletons and control strategies. At the end of the work, it is described the impact that exoskeleton versatility had on field testing.