Physical Human Machine Research Articles

The Influence of Circular Physical Human–Machine Interfaces of Three Shoulder Exoskeletons on Tissue Oxygenation

Occupational shoulder exoskeletons need to provide meaningful torques to achieve the desired support, thereby high pressures can occur within the physical human–machine interface (pHMI) of exoskeletons that may lead to discomfort, pain, or soft tissue injuries. This pilot study investigates the effects of occurring circumferential pressures within the pHMI in three different shoulder exoskeletons on the tissue oxygenation underneath the interfaces in resting position and dynamic use of the exoskeletons in 12 healthy subjects using near-infrared spectroscopy. Similar to standard Vascular Occlusion Tests, the tissue oxygen decreases while wearing the exoskeletons at rest (−2.1 (1.4) %/min). Dynamic use of the exoskeleton enhances the decrease in tissue oxygen (−7.3 (4.1) %/min) significantly and leads to greater resaturation after reopening the interface compared to resting position. This can be a sign of restricted blood supply to the upper extremity while wearing the exoskeleton. The shape and width of the circular interfaces showed no effect on the tissue oxygenation during use. Tissue oxygenation can be established as an additional safety criterion of exoskeletal pHMIs. The design of pHMI of shoulder exoskeletons should be reconsidered, e.g., in terms of open structures or the elasticity of closure straps to avoid occlusion effects.

Optimizing Exoskeleton Design with Evolutionary Computation: An Intensive Survey

Exoskeleton devices are designed for applications such as rehabilitation, assistance, and haptics. Due to the nature of physical human–machine interaction, designing and operating these devices is quite challenging. Optimization methods lessen the severity of these challenges and help designers develop the device they need. In this paper, we present an extensive and systematic literature search on the optimization methods used for the mechanical design of exoskeletons. We completed the search in the IEEE, ACM, and MDPI databases between 2017 and 2023 using the keywords “exoskeleton”, “design”, and “optimization”. We categorized our findings in terms of which limb (i.e., hand, wrist, arm, or leg) and application (assistive, rehabilitation, or haptic) the exoskeleton was designed for, the optimization metrics (force transmission, workspace, size, and adjustability/calibration), and the optimization method (categorized as evolutionary computation or non-evolutionary computation methods). We discuss our observations with respect to how the optimization methods have been implemented based on our findings. We conclude our paper with suggestions for future research.

Human Control Model Estimation in Physical Human-Machine Interaction: A Survey.

The study of human–machine interaction as a unique control system was one of the first research interests in the engineering field, with almost a century having passed since the first works appeared in this area. At the same time, it is a crucial aspect of the most recent technological developments made in application fields such as collaborative robotics and artificial intelligence. Learning the processes and dynamics underlying human control strategies when interacting with controlled elements or objects of a different nature has been the subject of research in neuroscience, aerospace, robotics, and artificial intelligence. The cross-domain nature of this field of study can cause difficulties in finding a guiding line that links motor control theory, modelling approaches in physiological control systems, and identifying human–machine general control models in manipulative tasks. The discussed models have varying levels of complexity, from the first quasi-linear model in the frequency domain to the successive optimal control model. These models include detailed descriptions of physiologic subsystems and biomechanics. The motivation behind this work is to provide a complete view of the linear models that could be easily handled both in the time domain and in the frequency domain by using a well-established methodology in the classical linear systems and control theory.

Low-Cost Force Sensors Embedded in Physical Human-Machine Interfaces: Concept, Exemplary Realization on Upper-Body Exoskeleton, and Validation.

In modern times, the collaboration between humans and machines increasingly rises, combining their respective benefits. The direct physical support causes interaction forces in human–machine interfaces, whereas their form determines both the effectiveness and comfort of the collaboration. However, their correct detection requires various sensor characteristics and remains challenging. Thus, this paper presents a developed low-cost sensor pad working with a silicone capsule and a piezoresistive pressure sensor. Its measurement accuracy is validated in both an isolated testing environment and a laboratory study with four test subjects (gender-balanced), and an application integrated in interfaces of an active upper-body exoskeleton. In the material-testing machine, it becomes apparent that the sensor pad generally features the capability of reliably determining normal forces on its surface until a certain threshold. This is also proven in the real application, where the measurement data of three sensor pads spatially embedded in the exoskeletal interface are compared to the data of an installed multi-axis load cell and a high-resolution flexible pressure map. Here, the consideration of three sensor pads potentially enables detection of exoskeletal support on the upper arm as well as “poor” fit conditions such as uneven pressure distributions that recommend immediate system adjustments for ergonomic improvements.

Summary of the Research on the Interaction of Tightly Coupled Physical Human-machine System

Summary of the Research on the Interaction of Tightly Coupled Physical Human-machine System

Effect of External Force on Agency in Physical Human-Machine Interaction.

In the advent of intelligent robotic tools for physically assisting humans, user experience, and intuitiveness in particular have become important features for control designs. However, existing works predominantly focus on performance-related measures for evaluating control systems as the subjective experience of a user by large cannot be directly observed. In this study, we therefore focus on agency-related interactions between control and embodiment in the context of physical human-machine interaction. By applying an intentional binding paradigm in a virtual, machine-assisted reaching task, we evaluate how the sense of agency of able-bodied humans is modulated by assistive force characteristics of a physically coupled device. In addition to measuring how assistive force profiles influence the sense of agency with intentional binding, we analyzed the sense of agency using a questionnaire. Remarkably, our participants reported to experience stronger agency when being appropriately assisted, although they contributed less to the control task. This is substantiated by the overall consistency of intentional binding results and the self-reported sense of agency. Our results confirm the fundamental feasibility of the sense of agency to objectively evaluate the quality of human-in-the-loop control for assistive technologies. While the underlying mechanisms causing the perceptual bias observed in the intentional binding paradigm are still to be understood, we believe that this study distinctly contributes to demonstrating how the sense of agency characterizes intuitiveness of assistance in physical human-machine interaction.

Admittance Control of a Teleoperated Motorized Functional Electrical Stimulation Rehabilitation Cycle

Abstract A bilateral teleoperated rehabilitation cycling system is developed for people with movement impairments due to various neurological disorders. A master hand-cycling device is used by the operator to set the desired position and cadence of a lower-body functional electrical stimulation (FES) controlled and motor assisted recumbent cycle. The master device also uses kinematic haptic feedback to reflect the lower-body cycle’s dynamic response to the operator. To accommodate for the unknown nonlinear dynamics inherent to physical human machine interaction (pHMI), admittance controllers were developed to indirectly track desired interaction torques for both the haptic feedback device and the lower-body cycle. A robust position and cadence controller, which is only active within the regions of the crank cycle where FES produces sufficient torque values, was used to determine the FES intensity. A Lyapunov analysis is used to prove the robust FES controller yields global exponential tracking to the desired position and cadence set by the master device within FES stimulation regions. Outside of the FES regions, the admittance controllers at the hands and legs work in conjunction to produce desired performance. Both admittance controllers were analyzed for the entire crank cycle, and found to be input/output strictly passive and globally exponentially stable in the absence of human effort, despite the uncertain nonlinear dynamics.

Attribution Biases and Trust Development in Physical Human-Machine Coordination: Blaming Yourself, Your Partner or an Unexpected Event

Reading partners’ actions correctly is essential for successful coordination, but interpretation does not always reflect reality. Attribution biases, such as self-serving and correspondence biases, lead people to misinterpret their partners’ actions and falsely assign blame after a surprise, or unexpected event. These biases further influence people’s trust in their partners, including machine partners (Muir, 1987; Madhavan & Wiegmann, 2004). Advances in robotics have allowed for robots to partner with people at work and be treated socially (Young, Hawkins, Sharlin & Igarashi, 2009). However, these advances may interfere with a person’s appropriate calibration of trust in robots (Parasuraman & Miller, 2004). A better understanding of attribution biases in the wake of an unexpected event may shed light on how trust develops in a robot partner. This study was built on a human coordination example to serve as a reference for future human-robot interactions. We posit that attribution biases lead people to blame their partner after experiencing a negative performance outcome, thus lowering their trust in the partner. Sixty participants (30 pairs) were tasked to coordinate with an unfamiliar human partner, to lift a 17.5 lb. box containing a 200ml cup of water filled to the brim, from the floor to a table, as quickly as possible without spilling water. Before the task, participants were told that the pair with the best performance would be rewarded; however, all pairs were told they did not achieve this. Participant pairs were randomly assigned to a surprise condition during which they heard a 250 Hz warning tone, or a baseline condition with no warning tone. Participants in both conditions were told to pause the task as quickly as possible if the warning tone was present. It was unknown to participants when or if a warning tone would occur. To assess participants’ trust in their partner, Muir’s (1987) trust questionnaire was administered twice, once after introducing the task to participants, and again after the coordination task was completed. To capture blame assignment, a scale based on Kim and Hinds (2006) was administered after participants were told they did not achieve the best performance. Results indicate participants were less likely to blame their partners for the negative outcome, compared to blaming themselves or the warning tone itself (in the surprise condition). Next, surprisingly, in the surprise condition, instead of experiencing a decrease of trust in a partner after the negative outcome, there was a significant increase in trust in their partners. No significant difference in trust was found in the baseline condition. Finally, results also indicate that initial trust in a partner is a significant predictor for how people assign blame. In general, the effects of attribution biases were not observed in the present study. Friendliness may be a factor in people’s assignment of blame; although participants were unfamiliar with one another, all participants were students at the same university. Second, shared experience during the surprise condition, including the chance to assess their partner’s behaviors in response to the warning tone, may have been a catalyst for increased trust in a partner. It is important to note that although physical differences between participants were not evaluated in this study, height may be a potential confounding factor in this task. These findings enlighten our understanding of physical human-robot coordination scenarios and trust in a partner.

Quantitative Modeling and Analysis of Reliance in Physical Human–Machine Coordination

Abstract Smooth and efficient human–machine coordination in joint physical tasks may be realized through greater sensing and prediction of a human partner's intention to apply force to an object. In this paper, we define compliance and reliance in the context of physical human–machine coordination (pHMC) to characterize human responses in a joint object transport task. We apply an optimization framework to explain human intention and behavior. The weighting factor in the optimization problem, lambda (λ), is presented as a person's reliance on the machine in a joint physical task with varying constraints. We demonstrate that with an estimated λ, the intended two-dimensional motion of a person's trajectory can be captured. We also found a relationship between λ and trust while participants performed a familiar task with no distraction. This finding suggests a relationship between the psychological construct of trust and joint physical coordination. The extent to which λ may serve as an online measure of trust and reliance in a physical load sharing task requires further investigation under more complex task scenarios that involve greater degrees of vulnerability and uncertainty.

Stability and Control of a Rider–Bicycle System: Analysis and Experiments

We present stability and control analysis of a rider–bicycle system under human steering and body movements. The dynamic model of rider–bicycle interactions is first constructed to integrate the rider’s body movement with the moving bicycle platform. We then present human balance control strategies based on human riding experiments. The closed-loop system stability is analyzed and discussed. Quantitative influences of the bicycle physical parameters, the human control gains, and the time delays are also analyzed and discussed. Extensive experiments are conducted to validate the human control models and demonstrate human balance performance using the bikebot, an instrumented bicycle platform. The presented modeling and analysis results can be potentially used for further development of bicycle-assisted rehabilitation for postural balance patients. Note to Practitioners —Understanding human balance and control of bicycles is crucial for not only designing bicycle-based rehabilitation devices but also studying physical human–machine interactions for healthcare automation. This paper takes the rider and bicycle as an example of physical human–machine interactions to understand how human use their limbs and body movement to stabilize an unstable platform (i.e., bicycles). We develop an instrumented bicycle system, called bikebot, to conduct human riding experiments. Using experimental data, we build the dynamic models for human steering and leaning control actions. Using these actuation models, stability analyses are conducted for rider–bicycle interactions and then validated by experiments. We also obtain the stability results by perturbing human visual and sensorimotor feedback. These results reveal that the visual feedback and the time delay in sensorimotor feedback mechanism play critical roles in stabilizing the unstable bicycle platform.

A Soft Three-Axis Force Sensor Based on Radially Symmetric Pneumatic Chambers

In physical human–machine interaction applications, the accurate measurement of interactive forces between the human and the machine plays a significant role. Sensors for this purpose should not only be accurate and reliable, but also be soft enough to guarantee the safe and compliant interaction. In this aspect, pneumatic sensors with soft air chambers have often been utilized as a soft force sensor system. Although the pneumatic-based force sensor provides a good compliance and softness, however, it measures only a lumped force acting on the chamber, because the pneumatic sensor measures the overall pressure change in the air chamber. In this paper, a three-axis force sensor with three air chambers in a radially symmetric pattern is proposed for the measurement of multi-dimensional interaction forces with high softness and compliance. Each air chamber embeds a pneumatic sensor, and the pressure changes in the three air chambers are measured in order to distinguish the direction of the applied force. By decoupling the sensor signals from the three pneumatic sensors, the three-dimensional force components can be calculated accurately. Consequently, the proposed sensor is able to measure the three-dimensional forces while maintaining the great softness and compliance provided by the soft air chambers. The design, the fabrication method, and the verification of the proposed method are introduced in this paper.

Cyber Physical Systems and Human-Machine (Robot) Collaboration

Cyber Physical Systems and Human-Machine (Robot) Collaboration

Hierarchical Human Machine Interaction Learning for a Lower Extremity Augmentation Device

For several years considerable effort has been devoted to the study of human augmentation robots. Traditionally, the focus of exoskeleton system has always been on model-based control framework. It seeks to model the dynamic system from prior knowledge of the robot as well as the pilot. However, in lower extremity exoskeleton, the control method depends on not only the modelling accuracy but also the physical human–machine interaction changed from personal physical conditions. To address this problem, in this paper, we present a model-free incremental human–machine interaction learning methodology. In a higher level, the methodology can plan the motion of exoskeleton with the sequence of rhythmic movement primitives. In the lower level, the gain scheming is updated from the dynamic system based on a novel proposed learning algorithm efficient $$PI^2$$ -CMA-ES. Compared with $$PI^{BB}$$ , a particular feature is that it directly operates on the Cholesky decomposition of the covariance matrix, reducing the computational effort from $$O(n^3)$$ to $$O(n^2)$$ . To evaluate our proposed methodology, we not only demonstrate its applications on the single leg exoskeleton platform but also test on our lower extremity augmentation device. Experimental results show that the proposed methodology can minimize the interaction between the pilot and the exoskeleton compared with the traditional model-based control strategy.

Whole-Body Pose Estimation in Physical Rider–Bicycle Interactions With a Monocular Camera and Wearable Gyroscopes

Abstract Pose estimation of human–machine interactions such as bicycling plays an important role to understand and study human motor skills. In this paper, we report the development of a human whole-body pose estimation scheme with application to rider–bicycle interactions. The pose estimation scheme is built on the fusion of measurements of a monocular camera on the bicycle and a set of small wearable gyroscopes attached to the rider's upper- and lower-limbs and the trunk. A single feature point is collocated with each wearable gyroscope and also on the body segment link where the gyroscope is not attached. An extended Kalman filter (EKF) is designed to fuse the visual-inertial measurements to obtain the drift-free whole-body poses. The pose estimation design also incorporates a set of constraints from human anatomy and the physical rider–bicycle interactions. The performance of the estimation design is validated through ten subject riding experiments. The results illustrate that the maximum errors for all joint angle estimations by the proposed scheme are within 3 degs. The pose estimation scheme can be further extended and used in other types of physical human–machine interactions.

Rapid Fabrication of Soft, Multilayered Electronics for Wearable Biomonitoring

Soft integrated electronics are key components for emerging applications in wearable biomonitoring, soft co‐robotics, and physical human–machine interaction. They are composed of soft and elastically deformable circuits and sensors that are combined with packaged microelectronics for signal processing, power regulation, and communication. While promising, widespread use of soft wearable electronics is currently limited by the lack of robust fabrication techniques to rapidly, efficiently, and precisely assemble soft and rigid components into multilayered systems. Here, an efficient digital fabrication approach is presented to create highly customizable wearable electronics through rapid laser machining and adhesion controlled soft materials assembly. Well aligned, multilayered materials are created from 2D and 3D elements that stretch and bend while seamlessly integrating with rigid components such as microchip integrated circuits, discrete electrical components, and interconnects. These techniques are applied using commercially available materials and components and the fabrication of thin, lightweight, customized sensor skins is demonstrated in under an hour. These fully integrated wireless devices conformably bond to the hand and are successfully used for monitoring hand gesture, pulse rate, and blood oxygenation. These materials and methods enable custom wearable electronics while offering versatility in design and functionality for a variety of applications through material selection and construction.

Development of an Intelligent Pneumatic Cylinder for Distributed Physical Human–Machine Interaction

Pneumatic systems are well known for their advantages and simplicity, and have been applied in various applications. This paper presents the development and experimental evaluation of an intelligent pneumatic cylinder and its control system. The cylinder is designed to have an optical encoder, pressure sensor, valve and a Programmable System on a Chip (PSoC) as the central processing unit. The PSoC will handle I2C communication, input and output data from the analogue to digital converter, counter program and pulse width modulation (PWM) duty cycle. An application tool for a distributed physical human–machine interaction is proposed using an intelligent pneumatic cylinder. The system applied 36 links of the actuator to form an Intelligent Chair Tool (ICT). The control methodology presented contains an inner force loop and an outer position loop implemented using a unified control system driven by PWM to an on/off valve. In this research, four control approaches, i.e., position control, force control, compliance control and viscosity control, were constructed and experimented. The physical properties of various objects were also detected by the intelligent cylinder through the detecting function experiment. Finally, an emulation experiment using mass was carried out and the results clearly show the ability of the intelligent cylinder, and the control approaches towards realization of the future ICT application.