force-torque measurement

This study addresses the challenge of sensor failure under high mechanical accelerations in structural dynamics experiments and the consequent degradation of traditional optimal experimental design (OED) performance. To overcome these limitations, the authors propose a robust OED framework that directly yields discrete sensor configurations. The approach integrates a relaxation strategy with gradient-based optimization and incorporates a binary-inducing regularizer to circumvent post-processing rounding heuristics. Leveraging a finite element model, the method is evaluated using the log-determinant of the parameter covariance matrix and mean squared error as performance metrics. Numerical results demonstrate that the proposed framework significantly outperforms classical designs across various sensor failure scenarios and efficiently handles high-dimensional, computationally expensive sensor placement problems.

To address the urgent need for compact, lightweight six-axis force/torque sensors in legged robots, this work proposes a non-contact sensing architecture based on optocouplers—overcoming key limitations of conventional strain-gauge sensors, such as mechanical fragility and significant thermal drift. A differentiable multiphysics parametric–performance mapping model is developed to enable joint global optimization of sensitivity and measurement error. The resulting sensor weighs less than 85 g and features full-scale ranges of ±150 N and ±5 N·m. Calibration and dynamic testing demonstrate a static error of ≤1.2%, with excellent agreement between theoretical predictions and empirical measurements. Integrated successfully onto a quadrupedal robot, the sensor accurately captures high-frequency ground–foot interaction forces. This work establishes a new paradigm for high-precision, robust force perception in highly dynamic legged platforms.

This study addresses the dual challenges of low-quality demonstrations and inefficient data collection in robotic imitation learning. We propose an instrumented demonstration framework, exemplified on a liquid dispensing task: a pressure sensor is embedded within a squeeze bottle, and a PI closed-loop controller enables high-precision, automated demonstration generation. Compared to conventional human demonstrations, our approach improves policy performance in 78% of test scenarios. Experiments show that instrumented collection not only substantially increases the volume of high-fidelity demonstration data but also enables Transformer-based imitation policies trained on automated demonstrations to outperform those trained on human demonstrations—on average—across evaluation metrics. To our knowledge, this is the first empirical validation that sensor-augmented automated demonstration collection simultaneously enhances data quality, scalability, and downstream policy generalization. Our work establishes a scalable, sensor-instrumented data infrastructure paradigm for developing general-purpose robotic agents.

To address the challenge of static sensor placement in digital twins—where fixed configurations fail to adapt to dynamic physical system changes, thereby limiting online data assimilation and prediction accuracy—this paper proposes an adaptive sensor reorientation strategy based on deep reinforcement learning (DRL), specifically leveraging DQN and PPO algorithms. Sensor reconfiguration is formulated as a Markov decision process, integrated with digital twin state representation and a structural health monitoring simulation platform to enable closed-loop, online optimization of sensing policies. This work represents the first application of DRL to dynamic sensor reconfiguration in digital twins, overcoming the limitations of conventional static or offline placement strategies. Experimental validation on a cantilever plate under multiple operational conditions (healthy/damaged states) demonstrates that dynamic sensor repositioning significantly increases information yield from measurements, reduces digital twin prediction error by 32.7%, and markedly improves the reliability of decision support.

Traditional robot design and control are typically decoupled, leading to morphologies poorly aligned with task requirements. This paper proposes a simulation-driven co-optimization framework for morphology and control, breaking the conventional “design-then-control” paradigm to enable task-oriented, end-to-end joint search. Our method employs gradient-free optimization to simultaneously evolve structural parameters and controller policies within a URDF-based multi-task reinforcement learning simulation environment. Key contributions include: (1) demonstrating that controller retraining significantly improves performance, yielding an average gain of 37%; and (2) revealing an inverse correlation between morphological complexity and controller training budget—providing theoretical justification for structural simplification under resource constraints. We validate the framework across four public simulation benchmarks, showing that co-optimization consistently yields more compact, robust, and task-adapted robot morphologies compared to sequential approaches.

This study addresses the limitations of traditional human dynamics analysis, which relies on contact-based force/torque sensors and controlled environments, rendering it unsuitable for non-contact scenarios. To overcome this, the authors propose an optics-mechanics integrated framework for simultaneous kinematic and dynamic estimation. By formulating a constrained multibody dynamics model, the method leverages vision-based kinematic measurements as non-contact inputs and employs a genetic algorithm to optimize joint torque identification. This approach represents the first demonstration of non-contact dynamic parameter estimation for multibody systems using only visual data and a mechanical model, eliminating dependence on force sensors. Experimental validation on an air-bearing platform shows a mean absolute error of 0.46 Nm in wrist joint torque estimation and a forward-predicted angular velocity error as low as 0.006 rad/s.

This study addresses the challenge of quantifying pitch-roll moment coupling in sub-gram flapping-wing aerial robots, which has been hindered by the absence of high-sensitivity dual-axis torque sensors. The authors propose a micromachined gimbal mechanism that enables, for the first time, high-precision simultaneous measurement of pitch moment, roll moment, and thrust. Integrated with a piezoelectric-driven flapping-wing platform and analyzed using linear regression and cross-correlation methods, experimental results demonstrate excellent linearity with coefficients of determination (R²) of 0.95 and 0.98 for pitch and roll moments, respectively. Crucially, the cross-coupling coefficients are nearly zero, and thrust fluctuations remain within 5.8% of the mean, confirming that pitch and roll moments can be approximately controlled independently. These findings provide critical experimental validation for decoupled control strategies in miniature flapping-wing vehicles.

Traditional engineering control systems struggle to replicate the robustness and agility exhibited by biological organisms in complex environments, primarily due to the neglect of active perception as a core component of task-level control. This work proposes that active perception serves not merely to reduce perceptual uncertainty but is intrinsically integrated into control at the task level. The study introduces a novel dual-mode “explore-exploit” control strategy that synergistically combines adaptive sensing, sensorimotor coupling, and dynamic behavioral mode switching. Through control-theoretic modeling, behavioral dynamical analysis, and biological empirical validation, the research elucidates the emergent mechanisms underlying biological active perception and establishes a new paradigm for enhancing perceptual and control capabilities in robotic systems.

This study addresses the absence of a quantitative framework for characterizing information propagation and processing capabilities in solid structures, which has hindered the unified design of mechanical functionality and information handling. For the first time, elastic solids are conceptualized as information encoders, integrating information theory with continuum mechanics to establish a quantitative metric for information transfer from external loads to discrete sensors. The work elucidates how geometry and architected materials govern information transmission pathways and efficiency. By linking classical mechanical phenomena—such as Saint-Venant’s principle and principal stress trajectories—to information-theoretic constructs, the authors propose benchmark tasks and evaluation metrics for mechanical intelligence, enabling the active design of information on/off states. This approach provides a foundational theoretical framework and a novel design paradigm for intelligent structural materials.

This study addresses the significant uncertainties in biomechanics arising from inter-individual variability and noisy experimental data by proposing a unified framework grounded in Bayesian probability theory. The approach systematically integrates forward uncertainty propagation and inverse parameter inference within a single coherent paradigm. It seamlessly combines input uncertainty characterization, data-driven surrogate modeling, model selection criteria, and information-theoretic optimal experimental design, while naturally linking global sensitivity analysis with spatially correlated random field priors. As the first mechanics-oriented uncertainty quantification methodology to adopt Bayesian inference as a unifying principle, this work delivers a theoretically consistent, computationally efficient, and robust toolkit for both computational and experimental mechanics, substantially enhancing the accuracy of parameter calibration and the reliability of predictive outcomes.