Shaping high-performance wearable robots for human motor and sensory reconstruction and enhancement

Abstract

Most wearable robots such as exoskeletons and prostheses can operate with dexterity, while wearers do not perceive them as part of their bodies. In this perspective, we contend that integrating environmental, physiological, and physical information through multi-modal fusion, incorporating human-in-the-loop control, utilizing neuromuscular interface, employing flexible electronics, and acquiring and processing human-robot information with biomechatronic chips, should all be leveraged towards building the next generation of wearable robots. These technologies could improve the embodiment of wearable robots. With optimizations in mechanical structure and clinical training, the next generation of wearable robots should better facilitate human motor and sensory reconstruction and enhancement.

Fig. 1. Breakthrough technologies for high-performance wearable robots.

Various technologies of wearable robots for motor and sensory enhancement and reconstruction are depicted. Biomechatronic chips could serve as the central unit for information acquisition and processing, generating control commands. Neuromuscular interface and flexible electronics are enabling technologies for sensing human intention, which could be fused together with multi-modal fusion. They also provide sensory feedback that transfers information from the robot to the human to improve the feeling of agency. Human-in-the-loop control integrates humans into the control loop of wearable robots, taking human reactions into account during the training process. (In the figure, arrows represent the information flow and dashed boxes represent technologies that form the key components of embodiment).

Fig. 2. Multi-modal information fusion for wearable robots.

Multi-modal signals were captured and extracted features, then fused with diverse fusion methods such as transient fusion and sequential fusion. The fusion result was used to recognize human intentions and account for personal differences to achieve individuation.

Fig. 3. Human-in-the-loop control.

In the process of using wearable robots, human biomechanical or physiological responses could be evaluated, and task performance could be compared over time. By using optimization methods, the wearable robot’s control strategy could be iteratively updated to optimization using effect.

Fig. 4. The bidirectional neuromuscular interface.

The sensing interface captures signals on the efferent nerve pathway (orange arrow downwards), which could be used to sense human intention, for example, ECoG, EEG, electroneurogram (ENG), implant EMG, and surface EMG. The feedback interface stimulates the afferent nerve pathway (indicated by the upward red arrow), which could be used to convey information to humans, for example, in non-invasive ways such as haptic feedback and surface electrical stimulation, in implantable ways like agonist-antagonist myoneural, implant electrode, and targeted sensory reinnervation.

Fig. 5. Flexible electronics for sensing and feedback.

It could be conformal with human skin or internal organs to capture signals or deliver stimulations. a Flexible electronics for sensing surface EMG. b Flexible electronics for sensing EEG. c Flexible electronics for sensing ECog. d Flexible electronics for sensing and feedback. e Flexible electronics for sensing invasive EMG. f Flexible electronics for sensing hand movement. g Flexible electronics for providing electrotactile. h Flexible electronics for nerve stimulation. i Flexible electronics for providing haptic feedback.

Fig. 6. Biomechatronic chips for wearable robots.

The signal acquisition part intakes multi-modal information, ADC prepares processable data, and neuromorphic computing handles on-chip neural network calculation. The biomechatronic chip serves as the central control unit for the wearable robot.