No abstract

Many applications for microengineered devices can be envisaged for actuators capable of doing work or transferring power. Millimetre order turbines are considered in this study for the development of torque and the possibilities for the delivery of work. A prototype microturbine, with overall thickness of less than a millimetre, was studied for its torque capabilities. The initial prototype was realized using precision mechanics although implementation of the turbine is planned using microengineering techniques. A viscous braking method was developed to measure the shaft torque of the turbine, demonstrating shaft coupling and the possibilities for power transfer. In order to validate the viscous braking method for torque measurement, a mechanical friction brake (dynamometer) was developed to compare the measurements obtained for a miniature electric motor of known characteristics. The results from this series of calibration experiments were then used to evaluate the performance of a microturbine prototype. The dynamometer torque measurements were found to closely agree with the manufacture's stated stall torque for the miniature motor of 1.8*10-4 N m. The viscous brake torque measurements were found to underestimate the motor torque by around 20% with slight variation related to the angular velocity of the shaft. Shaft torque measurements for the prototype microturbine were possible using the viscous brake but not the dynamometer. It was felt that 10-5 N m represented the lower limit for the dynamometer torque measurement while the viscous brake could address torques down to 10-8 N m. The fluid brake produced measurements of torque in the range of 10-7 N m for the microturbine. At this level only an order of magnitude accuracy is claimed because of some uncertainties with the fluid model used for the viscous brake torque calculation. The shaft torque range for the viscous brake was from 10-4 N m down to 10-8 N m; this might be extended by optimizing the fluid model.

No abstract

Automatic Speech Recognition (ASR) systems are increasingly powerful and more accurate, but also more numerous with several options existing currently as a service (e.g. Google, IBM, and Microsoft). Currently the most stringent standards for such systems are set within the context of their use in, and for, Conversational AI technology. These systems are expected to operate incrementally in real-time, be responsive, stable, and robust to the pervasive yet peculiar characteristics of conversational speech such as disfluencies and overlaps. In this paper we evaluate the most popular of such systems with metrics and experiments designed with these standards in mind. We also evaluate the speaker diarization (SD) capabilities of the same systems which will be particularly important for dialogue systems designed to handle multi-party interaction. We found that Microsoft has the leading incremental ASR system which preserves disfluent materials and IBM has the leading incremental SD system in addition to the ASR that is most robust to speech overlaps. Google strikes a balance between the two but none of these systems are yet suitable to reliably handle natural spontaneous conversations in real-time.

No abstract