Degradation evaluation of microelectromechanical thermal actuators

Metal based thermal microactuators normally have lower operation temperatures than those of Si-based ones; hence they have great potential for applications. However, metal-based thermal actuators easily suffer from degradation such as plastic deformation. In this study, planar thermal actuators were made by a single mask process using electroplated nickel as the active material, and their thermal degradation has been studied. Electrical tests show that the Ni-based thermal actuators deliver a maximum displacement of ∼20μm at an average temperature of ∼420°C, much lower than that of Si-based microactuators. However, the displacement strongly depends on the frequency and peak voltage of the pulse applied. Back bending was clearly observed at a maximum temperature as low as 240°C. Both forward and backward displacements increase with increasing the temperature up to ∼450°C, and then decreases with power. Scanning electron microscopy observation clearly showed that Ni structure deforms and reflows at power above 50mW. The compressive stress is believed to be responsible for Ni piling-up (creep), while the tensile stress upon removing the pulse current is responsible for necking at the hottest section of the device. Energy dispersive X-ray diffraction analysis revealed severe oxidation of the Ni-structure induced by Joule-heating of the current.

Enzyme-free glucose biosensor based on low density CNT forest grown directly on a Si/SiO2 substrate

A highly sensitive nonenzymatic amperometric glucose sensor was fabricated by using Ni nanoparticles homogeneously dispersed within and on the top of a vertically aligned CNT forest (CNT/Ni nanocomposite sensor), which was directly grown on a Si/SiO2 substrate. The surface morphology and elemental analysis were characterized using scanning electron microscopy and energy dispersive spectroscopy, respectively. Cyclic voltammetry and chronoamperometry were used to evaluate the catalytic activities of CNT/Ni electrode. The CNT/Ni nanocomposite sensor exhibited a great enhancement of anodic peak current after adding 5 mM glucose in alkaline solution. The sensor can also be applied to the quantification of glucose content with a linear range covering from 5 μM to 7 mM, a high sensitivity of 1433 μA mM-1 cm-2, and a low detection limit of 2 μM. The CNT/Ni nanocomposite sensor exhibits good reproducibility and long-term stability, moreover, it was also relatively insensitive to commonly interfering species, such as uric acid, ascorbic acid, acetaminophen, sucrose and d-fructose. © 2013 Elsevier B.V.

Evaluating the impact of V2G services on the degradation of batteries in PHEV and EV

Many researchers and industry observers claim that electric vehicles (EV) and plug-in hybrid electric vehicles (PHEV) could provide vehicle-to-grid (V2G) bulk energy and ancillary services to an electricity network. This work quantified the impact on various battery characteristics whilst providing such services. The sensitivity of the impact of V2G services on battery degradation was assessed for EV and PHEV for different battery capacities, charging regimes, and battery depth of discharge. Battery degradation was found to be most dependent on energy throughput for both the EV and PHEV powertrains, but was most sensitive to charging regime (for EVs) and battery capacity (for PHEVs). When providing ancillary services, battery degradation in both powertrains was most sensitive to individual vehicle battery depth of discharge. Degradation arising from both bulk energy and ancillary services could be minimised by reducing the battery capacity of the vehicle, restricting the number of hours connected and reducing the depth of discharge of each vehicle for ancillary services. Regardless, best case minimum impacts of providing V2G services are severe such as to require multiple battery pack replacements over the vehicle lifetime. © 2013 Elsevier Ltd.

High contrast holograms using nanotube forest

Nanotube forest behaves as highly absorbent material when they are randomly placed in sub-wavelength scales. Furthermore, it is possible to create diffractive structures when these bulks are patterned in a substrate. Here, we introduce an alternative to fabricate intensity holograms by patterning fringes of nanotube forest on a substrate. The result is an efficient intensity hologram that is not restricted to sub-wavelength patterning. Both the theoretical and experimental analysis was performed with good agreement. The produced holograms show a uniform behaviour throughout the visible spectra. © 2013 AIP Publishing LLC.

Unconstrained monocular 3D human pose estimation by action detection and cross-modality regression forest

This work addresses the challenging problem of unconstrained 3D human pose estimation (HPE) from a novel perspective. Existing approaches struggle to operate in realistic applications, mainly due to their scene-dependent priors, such as background segmentation and multi-camera network, which restrict their use in unconstrained environments. We therfore present a framework which applies action detection and 2D pose estimation techniques to infer 3D poses in an unconstrained video. Action detection offers spatiotemporal priors to 3D human pose estimation by both recognising and localising actions in space-time. Instead of holistic features, e.g. silhouettes, we leverage the flexibility of deformable part model to detect 2D body parts as a feature to estimate 3D poses. A new unconstrained pose dataset has been collected to justify the feasibility of our method, which demonstrated promising results, significantly outperforming the relevant state-of-the-arts. © 2013 IEEE.

The Random Forest Kernel and other kernels for big data from random partitions

We present Random Partition Kernels, a new class of kernels derived by demonstrating a natural connection between random partitions of objects and kernels between those objects. We show how the construction can be used to create kernels from methods that would not normally be viewed as random partitions, such as Random Forest. To demonstrate the potential of this method, we propose two new kernels, the Random Forest Kernel and the Fast Cluster Kernel, and show that these kernels consistently outperform standard kernels on problems involving real-world datasets. Finally, we show how the form of these kernels lend themselves to a natural approximation that is appropriate for certain big data problems, allowing $O(N)$ inference in methods such as Gaussian Processes, Support Vector Machines and Kernel PCA.