FG Mikro- und Nanosysteme
Refine
Year of publication
Document Type
Way of publication
- Open Access (13)
Keywords
Institute
Research profile line
Despite significant advances in telecommunications technologies, as well as the use of artificial intelligence in 5G and the anticipated reliance of 6G on trustworthy artificial intelligence, there are still insufficiently automated, inefficient, and unreliable areas of wireless network operation and maintenance that rely on human expert decisions. In this article, we point out the gap in automated performance monitoring and prediction as a factor preventing the design of autonomous mobile network management systems. In addition, we systematize the basics of modeling KPIs measured in a mobile network and conditioned by their configuration and environmental states. KPI modeling recommendations summarized for different network levels were demonstrated for two applications using datasets collected from operator networks, that is, quantifying the performance change in a population of 350 5G cells with unique configuration settings updates, and predicting the performance profile for an unknown configuration in a model trained with historical data associated with different configurations applied to a population of network cells.
With the advent of advanced transmission technologies, powerful neural-network architectures, and the availability of cheap sensing solutions, new paradigms for monitoring, controlling, and optimizing any automated system are appearing. While traditionally, sensors simply provided visual or acoustic feedback to human operators of such systems, nowadays, autonomous systems more and more rely on the automatic sensing and processing of such data to react and adapt their control mechanisms. In such complex control systems, sensor networks play a vital role. In this paper, we assess the state of the art in the three main components used within sensor networks: smart sensing elements, transmission technologies, and processing technologies. Concerning the latter, more and more artificial intelligence-based methods are developed and applied turning smart sensor networks into intelligent or even cognitive sensor networks. For the development of such sensing systems, a high variety of performance metrics could be important to the designer. This paper discusses the vast design space for intelligent sensor networks and provides guidance and directions on how to construct future intelligent sensor networks efficiently.
A novel concept of an acoustic flowmeter, based on single-mode waveguides, is proposed, implemented, and analysed in this work. Instead of transmitting a pulse diagonally across the duct's cross-section, this device operates with two ducts that operate simultaneously as pipes and as waveguides. Below the frequency threshold for single-mode propagation, acoustic waves are forced to traverse the waveguides with a plane front, precluding the possibility of beam drifting, inner reflections, and spreading losses. This enables the designer to flexibly increase the sound path and perform a highly sensitive measurement of the flow velocity and speed of sound, even if the excitation frequency is required to be kept below a relatively low value. A device based on this principle was constructed and tested for flow measurements in air. It consists of two waveguides of a circular cross-section (5 mmdiameter) coupled to electroacoustic transducers for the transmission of a wideband chirp (9.8–18.2 kHz). Usage of a wideband signal was possible due to the combined frequency response of a special kind of micromachined ultrasound transducer (MUT) and a commercial micro-electromechanical system (MEMS) microphone. The constructed flowmeter was capable of measuring flow velocities up until the transition to turbulent flow at 16 Lmin-1with a resolution of 0.3 Lmin-1, and it also detected changes of less than 0.2 ms-1in the speed of sound. This topology for flow measurement could prove advantageous for applications where gases of variable composition are conducted in ducts of diameters in the millimetre range.
Mobile spectral analysis has gathered a lot of interest throughout the last years. It enables the determination of object composition, material properties or other application relevant information directly on-site. This could be very interesting in agriculture, food business, environmental monitoring, medical or other fields. Autonomous operated systems could be handheld or mounted to robots as well as unmanned aerial vehicles (UAVs). The components have become compact and affordable so far. The evaluation of spectral data, for example in the near infrared range, requires stable calibration to extract the relevant information to extract concentration or composition information. Complex chemometric models and broad data bases are necessary. These are valid for selected devices only, transfer to other equipment is difficult. Furthermore, changes like aging of the light source must be calibrated from time to time. In close future artificial intelligence (AI) will help to change this situation by providing automated calibration and compensation routines.
We designed a MEMS microvalve based on the nanoscopic electrostatic drive (NED) technology (Nat Commun 6:10078, 2015). NED actuators, electrostatically controlled bending beams, are implemented in a clamped-clamped configuration. A normally open plunger valve was designed and characterized. The device is manufactured from silicon. Gas flow rates of up to 37 SCCM can be proportionally controlled between 10% and 100%. A 10% leakage is always present at low backpressures (< 10 kPa) and increases to roughly 20% at 75 kPa backpressure. The structure has been tested up to backpressures of 300 kPa without damage to the structures, but the leakage increases to over 95%. Our unprecedented microvalve concept shows that it is possible to manufacture all-silicon MEMS microvalves with proportional control of the flow rate. The presented work is a proof of concept to test the capabilities of the NED technology for the use in microvalves. There are plans to decrease the leakage in future designs by introducing an additional sealing layer as well as manufacturing a shutter instead of a plunger design.
The implementation of the “power-to-gas” concept, where hydrogen and natural gas are blended and transported in the existing network, requires a quick, on-site method to monitor the content of hydrogen in the mixture. We evaluate a rapid characterisation of this mixture based on the measurement of the speed of sound, using micromachined ultrasonic transducers (MUTs). Two MUT-based prototypes were implemented to analyse a mixture of natural gas and hydrogen under controlled conditions. Changes in the hydrogen content below 2 mol % (in a mixture that was adjusted between 6 mol % and 16 mol %) were discriminated by both devices, including the uncertainty due to the temperature compensation and the time-delay estimation. The obtained values of the speed of sound were consistent with those calculated from independent, non-acoustic measurements performed with a gas chromatograph and a density sensor. An MUT-based flow meter is thus capable of reporting both gas intake and the molar fraction of hydrogen, provided that the source of natural gas is kept constant.
We present the design, fabrication, and characterization of an innovative silicon-based micropump with high potential for portable lab-on-chip (LoC) as well as point-of-care (PoC) applications. The actuators of the pump are electrostatic driven in-plane bending devices, which were presented earlier (Borcia et al. in Phys Rev Fluids 3(8): 084202, 2018. 10.1103/PhysRevFluids.3.084202; Uhlig et al. in Micromachines, 9(4), 2018. 10.3390/mi9040190). This paper presents the characterization results achieved with the micropump. The dielectric non-polar liquid Novec7100™ was used as a test liquid due to its adequate physical properties. When applying a periodic voltage of 130 V, a flow rate of up to 80 µL/min was detected. The counter pressure amounts up to 30 kPa and the correspondent fluidic power (volumetric flow rate times the counter pressure) was calculated to 10 µW. The pump contains passive flap valves at the inlet and outlet, which are based on a bending cantilever design. Depending on the application requirements, the micropump can be designed modularly to adjust the specific parameters by an adequate arrangement of pump base units. In this paper, the proof of principle is shown using a single base unit with different number of stacked NED-actuator beams, as well as the serial arrangement of base units. Both modular concepts target the increase of backpressure of the NED-micropump in an inherently different way compared to conventional membrane micropumps.