Unravelings provide a probabilistic representation of solutions of master equations and a method of computation of the density operator dynamics. The trajectories generated by unravelings may also be treated as real—...
详细信息
Effective movement primitives should be capable of encoding and generating a rich repertoire of trajectories conditioned on task-defining parameters such as vision or language inputs. While recent methods based on the...
详细信息
Effective movement primitives should be capable of encoding and generating a rich repertoire of trajectories conditioned on task-defining parameters such as vision or language inputs. While recent methods based on the motion manifold hypothesis, which assumes that a set of trajectories lies on a lower-dimensional nonlinear subspace, address challenges such as limited dataset size and the high dimensionality of trajectory data, they often struggle to capture complex task-motion dependencies, i.e., when motion distributions shift drastically with task variations. To address this, we introduce Motion Manifold Flow Primitives (MMFP), a framework that decouples the training of the motion manifold from task-conditioned distributions. Specifically, we employ flow matching models, state-of-the-art conditional deep generative models, to learn task-conditioned distributions in the latent coordinate space of the learned motion manifold. Experiments are conducted on language-guided trajectory generation tasks, where many-to-many text-motion correspondences introduce complex task-motion dependencies, highlighting MMFP's superiority over existing methods.
Matrix Lie groups are an important class of manifolds commonly used in control and robotics, and optimizing control policies on these manifolds is a fundamental problem. In this work, we propose a novel computationall...
详细信息
In dealing with emergencies, rational decisions are needed with limited time to reduce losses to death. Emergencies, which affect many public interests, require the participation of decision-makers with different pers...
In dealing with emergencies, rational decisions are needed with limited time to reduce losses to death. Emergencies, which affect many public interests, require the participation of decision-makers with different perspectives to mask the lack of knowledge and experience of one decision-maker or a smaller group. Large-scale group decision-making (LSGDM) has become an exciting research topic in the last decade. Previous researchers have widely discussed the development and application of the LSGDM model. Time constraints and rational solutions are the main challenges for LSGDM in emergencies. This literature study explores LSGDM research trends in emergencies and provides new insights and opportunities for determining future research.
The point-of-care testing device has acquired importance for clinical diagnostics because of technological advances in resource-constrained situations and a decentralized healthcare system. The dependency on the point...
详细信息
Iterative linear quadratic regulator (iLQR) has gained wide popularity in addressing trajectory optimization problems with nonlinear system models. However, as a model-based shooting method, it relies heavily on an ac...
This study investigates the free vibration behaviour of simply supported square laminated composite plates with a square cut-out and straight fibre orientation designs. The cut-out is placed at the centre of the plate...
详细信息
Electrohydrodynamic jet (e-jet) printing is a promising additive manufacturing technique that has demonstrated high-resolution fabrication capabilities. Various jetting regimes as a function of the input voltage signa...
Electrohydrodynamic jet (e-jet) printing is a promising additive manufacturing technique that has demonstrated high-resolution fabrication capabilities. Various jetting regimes as a function of the input voltage signal have been identified and studied. In DC printing, a transition has been observed from the continuous jet to the natural pulsation mode, which precipitates less predictable outcomes on the print rate and quality of the final pattern. This phenomenon is not well understood, and its behavior is yet to be characterized. Accordingly, the natural pulsation initiation (NPI) of 15 solutions with unique viscosities and conductivities was analyzed. Dimensionless parameters referencing material properties (Ohnesorge and ratio of capillary to charge relaxation time), applied voltage (electric capillary), and induced flow rate (Weber) are implemented to analyze the dynamics of this transition. Our findings indicate that NPI is generally more prevalent in inks with higher conductivities. Additionally, less viscous inks appeared to inhibit NPI, providing a larger design space for attaining stable behavior.
The development of green batteries has implications for many fields including sustainable robotics and edible electronics. Here we present GelBat, a biodegradable, digestible and rechargeable battery constructed from ...
详细信息
The development of green batteries has implications for many fields including sustainable robotics and edible electronics. Here we present GelBat, a biodegradable, digestible and rechargeable battery constructed from gelatin and activated carbon. The device utilises the water splitting reaction to produce a simple, sustainable Bacon fuel cell which can produce an output voltage of over 1V for 10 minutes, depending on the load resistance, with 10 minutes of charging and whose only byproduct is water. Electrochemical impedance spectroscopy, cyclic voltammetry and self discharge tests are carried out to characterize the behaviour of the battery. The system does not lose any efficiency with repeated recharging cycles and can be completely dissolved in a simulated gastric fluid within 20 minutes. The simplicity of this design combined with the bioresorbable materials demonstrates the potential of this work to help advance robotic research towards more sustainable untethered autonomous systems and edible robots.
Safety concerns during the operation of legged robots must be addressed to enable their widespread use. Machine learning-based control methods that use model-based constraints provide promising means to improve robot ...
详细信息
暂无评论