Exoskeletons can assist humans during squatting and the assistance has the potential to reduce the physical demands. Although several squat assistance methods are available, the effect of personalized assistance on physical effort has not been examined. We hypothesize that personalized assistance will reduce the physical effort of squatting. We developed a human-in-the-loop Bayesian optimization scheme to minimize the metabolic cost of squatting using a unilateral ankle exoskeleton. The optimization identified subject-specific assistance parameters for ascending and descending during squatting and took 15.8 min on average to converge. The subject-specific optimized condition reduced metabolic cost by 19.9% and rectus femoris muscle activity by 28.7% compared to the condition without the exoskeleton with a higher probability of improvement compared to a generic condition. In an additional study with two participants, the personalized condition presented higher metabolic cost reduction than the generic condition. These reductions illustrate the importance of personalized ankle assistance using an exoskeleton for squatting, a physically intensive activity, and suggest that such a method can be applied to minimize the physical effort of squatting. Future work can investigate the effect of personalized squat assistance on fatigue and the potential risk of injury. Index Terms-Squatting, human-in-the-loop, ankle-foot orthosis, and exoskeleton. I. INTRODUCTIONS QUATTING is a physically intensive and injury-prone task. Joint loading during squatting is higher than it is in other daily activities, such as walking [1], [2] and standing. Joint loads in squatting have increased contact stress and injuries in the tibiofemoral and patellofemoral joints [3]-[5].

Walking is an important activity that supports the health-related quality of life, and for those who need assistance, robotic devices are available to help. Recent progress in wearable robots has identified the importance of customizing the assistance provided by the robot to the individual, resulting in robot adaptation to the human. However, current implementations minimize the role of human adaptation to the robot, for example, by the users modifying their movements based on the provided robot assistance. This study investigated the effect of visual feedback to guide the users in adapting their movements in response to wearable robot assistance. The visual feedback helped the users reduce their metabolic cost of walking without any changes in robot assistance in a given time. In a case with the initially metabolic expensive (IMExp) exoskeleton condition, both training methods helped reduce the metabolic cost of walking. The results suggest that visual feedback training is helpful to use the exoskeleton for various conditions. Without feedback, the training is helpful only for the IMExp exoskeleton condition. This result suggests visual feedback training can be useful to facilitate the use of non-personalized, generic assistance, where the assistance is not tuned for each user, in a relatively short time.

Individuals with below-knee amputation (BKA) experience increased physical effort when walking, and the use of a robotic ankle-foot prosthesis (AFP) can reduce such effort. The walking effort could be further reduced if the robot is personalized to the wearer using human-in-the-loop (HIL) optimization of wearable robot parameters. The conventional physiological measurement, however, requires a long estimation time, hampering real-time optimization due to the limited experimental time budget. This study hypothesized that a function of foot contact force, the symmetric foot force-time integral (FFTI), could be used as a cost function for HIL optimization to rapidly estimate the physical effort of walking. We found that the new cost function presents a reasonable correlation with measured metabolic cost. When we employed the new cost function in HIL ankle-foot prosthesis stiffness parameter optimization, 8 individuals with simulated amputation reduced their metabolic cost of walking, greater than 15% (p < 0.02), compared to the weight-based and control-off conditions. The symmetry cost using the FFTI percentage was lower for the optimal condition, compared to all other conditions (p < 0.05). This study suggests that foot force-time integral symmetry using foot pressure sensors can be used as a cost function when optimizing a wearable robot parameter.

The foot center of pressure (COP) variability is an important indicator of balance, particularly relevant for rehabilitation and training using wearable lower limb exoskeletons. This study aimed to evaluate the effectiveness of our exoskeleton in assisting squatting motion using the COP variability as a metric. Six human subjects performed alternate squatting and standing movements while their foot pressure and COP trajectories were recorded using insole pressure sensors. The exercises were performed under three conditions: i) no device, ii) unpowered device, and iii) device with optimal stiffness. Results showed that the variability of the COP trajectory in the anterior-posterior direction of the foot during squatting tended to be lower for the optimal stiffness condition than the no device and unpowered device conditions, indicating the potential usefulness of the device in improving balance during squatting. This study has implications for human-inthe-loop optimization and balance control of the exoskeleton based on COP.

Squatting is an intensive activity routinely performed in the workplace to lift and lower loads. The effort to perform a squat can decrease using an exoskeleton that considers individual worker’s differences and assists them with a customized solution, namely, personalized assistance. Designing such an exoskeleton could be improved by understanding how the user’s muscle activity changes when assistance is provided. This study investigated the change in the muscle recruitment and activation pattern when personalized assistance was provided. The personalized assistance was provided by an ankle–foot exoskeleton during squatting and we compared its effect with that of the no-device and unpowered exoskeleton conditions using previously collected data. We identified four main muscle recruitment strategies across ten participants. One of the strategies mainly used quadriceps muscles, and the activation level corresponding to the strategy was reduced under exoskeleton assistance compared to the no-device and unpowered conditions. These quadriceps dominant synergy and rectus femoris activations showed reasonable correlations (r = 0.65, 0.59) to the metabolic cost of squatting. These results indicate that the assistance helped reduce quadriceps activation, and thus, the metabolic cost of squatting. These outcomes suggest that the muscle recruitment and activation patterns could be used to design an exoskeleton and training methods.