The integration of advanced sensor technologies has significantly propelled the dynamic development of robotics, thus inaugurating a new era in automation and artificial intelligence. Given the rapid advancements in robotics technology, its core area—robot control technology—has attracted increasing attention. Notably, sensors and sensor fusion technologies, which are considered essential for enhancing robot control technologies, have been widely and successfully applied in the field of robotics. Therefore, the integration of sensors and sensor fusion techniques with robot control technologies, which enables adaptation to various tasks in new situations, is emerging as a promising approach. This review seeks to delineate how sensors and sensor fusion technologies are combined with robot control technologies. It presents nine types of sensors used in robot control, discusses representative control methods, and summarizes their applications across various domains. Finally, this survey discusses existing challenges and potential future directions.

Surface electromyography- (sEMG-) based hand grasp force estimation plays an important role with a promising accuracy in a laboratory environment, yet hardly clinically applicable because of physiological changes and other factors. One of the critical factors is the muscle fatigue concomitant with daily activities which degrades the accuracy and reliability of force estimation from sEMG signals. Conventional qualitative measurements of muscle fatigue contribute to an improved force estimation model with limited progress. This paper proposes an easy-to-implement method to evaluate the muscle fatigue quantitatively and demonstrates that the proposed metrics can have a substantial impact on improving the performance of hand grasp force estimation. Specifically, the reduction in the maximal capacity to generate force is used as the metric of muscle fatigue in combination with a back-propagation neural network (BPNN) is adopted to build a sEMG-hand grasp force estimation model. Experiments are conducted in the three cases: (1) pooling training data from all muscle fatigue states with time-domain feature only, (2) employing frequency domain feature for expression of muscle fatigue information based on case 1, and 3) incorporating the quantitative metric of muscle fatigue value as an additional input for estimation model based on case 1. The results show that the degree of muscle fatigue and task intensity can be easily distinguished, and the additional input of muscle fatigue in BPNN greatly improves the performance of hand grasp force estimation, which is reflected by the 6.3797 increase in R² (coefficient of determination) value.

Lateral walking gait detection is necessary for the development of wearable devices applied to side stepping. To our knowledge, rarely work has been conducted to identify lateral walking gait by wearable sensors. Based on a hip exoskeleton, we presented a method for lateral walking gait phase detection only by two IMUs mounted on the shank. Experiments were conducted to detect narrow double support, swing of the leading leg, wide double support and swing of trailing leg phases of 12 healthy subjects walking at various speeds. The performance of four different algorithms including thresholding (THR), modified k-nearest neighbor named urban buildings indexing (UBI), random forest (RF) and neural networks (NN) were evaluated. The occupation of space resources of NN is the smallest besides THR. The total recognition accuracy (mean and SEM) of the RF-based, UBI-based, and NN-based systems was 97.07 ± 0.07% (off-line), 96.64 ± 0.16% (off-line) and 95.22 ± 0.60% (real-time), respectively. The recognition time of the models based on RF, UBI, NN, and THR was 13.3 ± 1.1 ms, 5.7 ± 0.5 ms, 2.6 ± 0.2 ms and 0.8 ± 0.2 ms, respectively. The recognition accuracy (cross-subjects) of RF-based, UBI-based and NN-based systems was 91.72 ± 0.42%, 89.63 ± 0.48% and 89.60 ± 0.43%, respectively. The results demonstrated that the proposed method can be applied to lateral walking gait detection.

To obtain accurate output torque of robot joint for torque control, a built-in torque sensor is designed in this paper. In order to enhance the performance of harmonic gear, it is urgent to deal with the key issues in terms of the errors of sensing and positioning, torque ripple interference. To handle these aforementioned key issues, four pairs of symmetrically distributed strain gauges are pasted on the bottom of the flexspline. Then, a precise positioning method based on laser marking is proposed to improve the positioning accuracy of the strain gauges. Moreover, Kalman filter is utilized to reduce torque ripple interference in the output signal of the strain gauges. Additionally, the designed device is evaluated by experimental tests which is conducted on prototype trial production. Synchronously, the relationship between the flexspline torque and the joint output torque is verified and the polynomial is used to represent the relationship of the measured torque for the built-in torque sensor and the standard torque sensor in the static and dynamic conditions. Finally, a friction model is presented and identified. Comprehensive comparative experiments is performed the performance of the build-in torque sensor in terms of estimation accuracy and control and it shows that the designed built-in torque sensor with friction compensation can be improved ten times in comparison with that of without friction compensation in terms of steady state mean error and mean value of measurement error. What is more, this designed build-in torque sensor holds tremendous potential for making the compact and light weight motion joint module, especially for cooperative robot and robotic supernumerary limb.

Legged robots are capable of achieving energy-efficiency by introducing elastic elements in parallel with the actuators during the stance phase. However, the parallel elastic actuation results in a coupling motion between the parallel spring and the flexion movement of the leg mechanism during the swing phase. This coupling hinders the dexterity of joint movement and increases energy consumption which degrades the energy-efficiency of legged robots. In this letter, we present an origami-based decoupling clutch for energy-efficient legged robots. The decoupling mechanism is realized by a Kresling-based clutch that switches the connection between the parallel spring and the robotic leg without additional control effort. During stance, the parallel spring and the leg are engaged to reduce the torque requirements. Furthermore, the leg can achieve unobstructed movement when the spring is disengaged during the swing phase. Subsequently, we design a single-leg prototype to validate the enhancement of energy-efficiency. Experimental results demonstrate that the robot with the proposed origami-based decoupling clutch can reduce the required knee flexion torque and energy consumption by 37.6% and 31%, respectively, of what is required for the robot without the clutch.