Mid Sweden University

Postural sway is a critical measure for evaluating postural control, and its analysis plays a vital role in preventing falls among the elderly. Typically, physiotherapists assess an individual's postural control using tests such as the Berg Balance Scale, Tinetti Test, and time up-and-go test. Sensor-based analysis is available based on devices such as force plates or inertial measurement units. Recently, machine learning methods have demonstrated promising results in the sensor-based analysis of postural control. However, these models are often complex, slow, and energy-intensive. To address these limitations, this study explores the design space of lightweight machine learning models deployable to microcontrollers to assess postural stability. We developed an artificial neural network (ANN) model and compare its performance to that of random forests, gaussian naive bayes, and extra tree classifiers. The models are trained using a sway dataset with varying input sizes and signal-to-noise ratios. The dataset comprises two feature vectors extracted from raw accelerometer data. The developed models are deployed to an ARM Cortex M4-based microcontroller, and their performance is evaluated and compared. We show that the ANN model has 99.03% accuracy, higher noise immunity, and the model performs better with a window size of one second with 590.96 us inference time. 

Acoustic emission (AE) is a widely used non-destructive test method in structural health monitoring applications to identify the damage type in the material. Usually, the analysis of the AE signal is done by using traditional parameter-based methods. Recently, machine learning methods showed promising results for the analysis of AE signals. However, these machine learning models are complex, slow, and consume significant amounts of energy. To address these limitations and to explore the trade-off between model complexity and the classification accuracy, this paper presents a lightweight artificial neural network model to classify damage types in concrete material using raw acoustic emission signals. The model consists of one hidden layer with four neurons and is trained on a public acoustic emission signal dataset. The created model is deployed to several microcontrollers and the performance of the model is evaluated and compared with a state-of-the-art machine learning model. The model achieves 98.4% accuracy on the test data with only 4019 parameters. In terms of evaluation metrics, the proposed tiny machine learning model outperforms previously proposed models 10 to 1000 times. The proposed model thus enables machine learning in real-time structural health monitoring applications. 

Condition monitoring devices in hydraulic systems that use batteries or require wired infrastructure have drawbacks that affect their installation, maintenance costs, and deployment flexibility. Energy harvesting technologies can serve as an alternative power supply for system loads, eliminating batteries and wiring requirements. Despite the interest in pressure fluctuation energy harvesters, few studies consider end-to-end implementations, especially for cases with lowamplitude pressure fluctuations. This generates a research gap regarding the practical amount of energy available to the load under these conditions, as well as interface circuit requirements and techniques for efficient energy conversion. In this paper, we present a self-powered sensor that integrates an energy harvester and a wireless sensing system. The energy harvester converts pressure fluctuations in hydraulic systems into electrical energy using an acoustic resonator, a piezoelectric stack, and an interface circuit. The prototype wireless sensor consists of an industrial pressure sensor and a low-power Bluetooth System-on-chip that samples and wirelessly transmits pressure data. We present a subsystem analysis and a full system implementation that considers hydraulic systems with pressure fluctuation amplitudes of less than 1 bar and frequencies of less than 300 Hz. The study examines the frequency response of the energy harvester, the performance of the interface circuit, and the advantages of using an active power improvement unit adapted for piezoelectric stacks. We show that the interface circuit used improves the performance of the energy harvester compared to previous similar studies, showing more power generation compared to the standard interface. Experimental measurements show that the self-powered sensor system can start up by harvesting energy from pressure fluctuations with amplitudes starting at 0.2 bar at 200 Hz. It can also sample and transmit sensor data at a rate of 100 Hz at 0.7 bar at 200 Hz. The system is implemented with off-the-shelf circuits. 

Pressure fluctuation energy harvesting devices are promising alternatives to power up wireless sensors in fluid power systems. In past studies, classical Helmholtz resonators have been used to enhance the energy harvesting capabilities of these harvesters. Nevertheless, for fluctuations with frequency components in the range of less than 1000 Hz, the design of compact resonators is difficult, mostly for their poor acoustic gain. This paper introduces a space-coiling resonator fabricated using 3D printing techniques. The proposed resonator can achieve a better acoustic gain bounded by a small bulk volume compared to a classic Helmholtz resonator, improving the energy harvesting capabilities of pressure fluctuation energy harvesters. The resonator is designed and evaluated using finite-element-method techniques and examined experimentally. Three space-coiling-resonators are designed, manufactured and compared to classic Helmholtz resonators for three frequencies: 280 Hz, 480 Hz and 920 Hz. This work displays the possibility of compact, high-performance pressure fluctuation energy harvesters and the advantages of the space-coiling printed resonators to enhance the harvesting performance.

Hydraulic pressure fluctuation energy harvesters are promising alternatives to power up wireless sensor nodes in hydraulic systems. The characterization of these harvesters under dynamic and band-limited pressure signals is imperative for the research and development of novel concepts. To generate and control these signals in a hydraulic medium, a versatile apparatus capable of reproducing pressure signals is proposed. In this paper, a comprehensive discussion of the design considerations for this apparatus and its performance is given. The suggested setup enables the investigation of devices tailored for the harvesting of energy in conventional hydraulic systems. To mimic these systems, static pressures can be tuned up to 300 bar, and the pressure amplitudes with a maximum of 28 Bar at 40 Hz and 0.5 bar at 1000 Hz can be generated. In addition, pressure signals found in commercial hydraulic systems can be reproduced with good accuracy. This apparatus proves to be an accessible, robust, and versatile experimental setup to create environments for the complete performance estimation of pressure fluctuation energy harvesters. 

Wireless sensor nodes in state of the art fluid power systems used in monitoring and maintenance prediction demand long lasting power sources that do not rely on batteries. Energy harvesting is a promising technology that can provide the required energy to power wireless sensors. Pressure fluctuation energy harvesters can be employed in conventional hydraulic systems to convert the acoustic pressure fluctuation to electrical power. Present studies have explored the overall efficiency of these devices while experimentally describing losses in piezoelectric and circuit interfaces, nevertheless there is no study on the fluid to mechanical force transmission efficiency. In this paper we investigate the pressure to force transmission rate of two types of fluid to mechanical interfaces: a flat metal plate and a conventional hydraulic piston. The interfaces are investigated in conditions similar to those found in conventional hydraulic systems. The study shows that flat plate exhibit good force transmission for low pressure applications with a constant rate across frequencies, while exhibiting a decrease in force transmission at higher pressures. On the other hand the piston exhibit a more robust pressure design, with a constant force transmission rate at all pressures but with a dampening of force at higher frequencies. It is shown that small differences in force transmission ratios can have a considerable impact on the power generation.

The need for wireless sensor networks that can run for long times without the need of battery replacement has risen the need for energy harvesters. Industrial environments have plenty of energy sources that can be harvested; pressure fluctuations are a high energy density source that can be harvested using piezoelectric devices. Present devices have introduced flat metallic plates as the main force transmission elements for hydraulic fluctuations energy harvesters. In this paper, we analyze the force transmission efficiency of flat plates when used as the primary fluid coupling interface in hydraulic energy harvesters. Previous work has been focused on the optimization of circuit matching and pressure ripple amplification. In this work, we offer a look into the efficiencies of flat plates in different configurations and pressure loads. The analysis shows that despite the reasonable force transmission efficiency of flat plates in low-pressure environments, the overall efficiency of hydraulic energy harvesters can be improved if instead of flat plates, conventional hydraulic actuators, such as piston cylinders, could be used. 

Wireless Sensor Networks appear as a technology, which provides the basisfor a broad field of applications, drawing interest in various areas. On theone hand, they appear to allow the next step in computer networks, buildinglarge collections of simple objects, exchanging information with respect totheir environment or their own state. On the other hand, their ability tosense and communicate without a fixed physical infrastructure makes theman attractive technology to be used for measurement systems.Although the interest inWireless Sensor Network research is increasing,and new concepts and applications are being demonstrated, several fundamentalissues remain unsolved. While many of these issues do not requireto be solved for proof-of-concept designs, they are important issues to beaddressed when referring to the long-term operation of these systems. Oneof these issues is the system’s lifetime, which relates to the lifetime of thenodes, upon which the system is composed.This thesis focuses on node lifetime extension based on energy management.While some constraints and results might hold true from a moregeneral perspective, the main application target involves environmental measurementsystems based onWireless Sensor Networks. Lifetime extensionpossibilities, which are the result of application characteristics, by (i) reducingenergy consumption and (ii) utilizing energy harvesting are to be presented.For energy consumption, we show howprecise task scheduling due to nodesynchronization, combined with methods such as duty cycling and powerdomains, can optimize the overall energy use. With reference to the energysupply, the focus lies on solar-based solutions with special attentionplaced on their feasibility at locations with limited solar radiation. Furtherdimensioning of these systems is addressed.It will be shown, that for the presented application scenarios, near-perpetualnode lifetime can be obtained. This is achieved by focusing on efficient resourceusage and by means of a carefully designed energy supply.

Technological development has through time changed and improved the way measurements are being performed. Starting from entirely mechanical designs, today's measurement instruments are electronic, computerized and, in many cases, connected. This has enabled a largely automated collection of physical quantities with high resolution and reliability. The recorded data may be used as the basis for decision making or may be utilized in closed-loop process control.

Wireless sensor networks have been demonstrated, at an early stage in their development, to be a useful measurement technology for environmental monitoring applications. Based on their independence from existing infrastructures, wireless sensor networks can be deployed in virtually any location and provide sensor samples in a spatial and temporal resolution, which otherwise would only be achievable at high cost or involve significant work by humans.The feasibility of the usage of wireless sensor networks in real-world applications, however, is only maintained if certain technological challenges are overcome. Amongst these challenges, are the limited lifetime of the distributed sensor nodes, and user interfaces, which allow for the technology to be utilized in an efficient manner. Contributions to the solution of these challenges have been the objective of this thesis.

After an analysis of the contributions wireless sensor networks can provide

to the application domain of environmental monitoring, and the introduction

to the restrictions, which are posed by a limited operational lifetime and low

system usability, these issues are addressed at the system level of sensor nodedevices.

The lifetime of sensor nodes, which is closely linked to the lifetime of the

complete wireless sensor network, is addressed with regards to the energy

efficiency of nodes, as well as the utilization of solar energy harvesting in

order to increase the available energy resources. With respect to energy

efficiency, an analysis has been performed of the contributions to the energy

consumption of environmental monitoring sensor nodes, which leads to the

desire to minimize the nodes' duty cycles and quiescent currents. A sensor

node design is presented, which features energy efficiency as a key attribute by utilizingmodern semiconductor architectures. Moreover, an argument for the usage of synchronization-based, contention-free communication is made

in order to reduce active communication periods and, thus, the duty cycle

of a sensor node. A synchronization method with its focus on low protocol

overhead is introduced as a basis for such communication forms. After an initial feasibility study in relation to using battery-less solar energy

harvesting architectures in locations with limited solar irradiation, multiple

architectural implementations are analyzed in a comparative manner.

Among these comparisons is an analysis of short-termenergy storage devices in the form of double-layer capacitors and thin-film batteries, which provide prolonged component lifetimes than those for conventional secondary batteries, but which can only buffer for short periods of time due to their limited energy capacity. In order to be able to dimension such energy harvesting systems with respect to the individual application constraints at hand, state of charge simulations are proposed. Amethod for such simulations is presented and demonstrated for the implementation of an energy harvester model on a component basis. While the modeling in this manner is time consuming, the model can predict the state of charge of the energy buffer in the architecture with a high level of accuracy. Finally, a method for the systematic evaluation of solar energy harvesting architectures is presented. The presented method can be summarized as a solar energy harvesting testbed, which utilizes configurable energy harvesting circuits in order to create a deploy-once-test-many type of system. The output results of this testbed can significantly improve

the efficiency of architecture comparisons and system modeling.

Contributions to the improvement of the usability of wireless sensor nodes

are made on two separate levels, namely, developer usability and end user

usability. A method for the programming of sensor nodes based on hierarchical finite state machines is presented, which improves the usability of software development by creating familiarity for technically experienced

users. Moreover, the utilization of finite state machine principles allows

for the software to be developed in a systematic andmodular manner. As

implemented applications typically require to be verified, which, in the environmental monitoring domain, usually results in outdoor deployments,

usability considerations for sensor nodes are presented, which can simplify

this process. Special attention has been paid in order for these improvements to be achieved with low overheads. While software development is a familiar concept for most system developers, this is not the case for the end users of these systems, who are typically domain experts. In order to allow for wireless sensor nodes to be operated

by domain experts, a method for the configuration of sensor nodes has been proposed.The method uses a combination of graphical specification of the node behavior and a configurable sensor node. Theevaluation of this method, which has been based on a proof-of-concept implementation, demonstrated that the performance can remain high, while end users, without technical experience, are enabled to configure sensor nodes without prior training.

In summary, the contributions, presented in this thesis, address system

lifetime and usability with regards to the sensor node level. The results have

led to the implementation of an energy efficient sensor node, which allows for the operation frombattery-less solar energy harvesting sources. Furthermore, support tools for the implementation of these nodes, both on the hardware and software level, have been proposed.

Nowadays, air pollution is monitored with accurate, but large-sized measurement stations, leading to an overall limited number of monitored locations. Combining these stations, with a higher number of less accurate stations can provide additional information, such as with regards to pollutant distributions. In this paper we present the design, implementation and initial results of such stations based on Wireless Sensor Network technology. For the implementation of the network purely off-the-shelf equipment was chosen, which allows us to analyze the current status of commercially available Wireless Sensor Network technology. While the system was fully implemented and demonstrated operationally, the experiences found during the project showed a limited matureness with regards to the off-the-shelf equipment and uncovered flaws in typical assumptions underlying Wireless Sensor Network research. © 2011 IEEE.

This paper addresses one large university initiative for educational development aimed at further developing educations and teacher competence with a focus on technology-enhanced and lifelong learning. The aim of the paper is to describe and problematize the design of an ongoing project for educational development, Higher Education and Digitalisation (HEaD). It focuses on identifying key components of an educational development project for technology enhanced learning as well as how such a project can be organized to sustain in regular university operations. The article discusses how a project for educational development can create over-time durable infrastructures, organization, policy and motivation for maintaining a continual educational development. In the first phase of the project, a model was developed for how competence development can be conducted sustainably. This model contains two perspectives: (1) an organizational perspective that focuses on the key partners to be involved; and (2) a process perspective that focuses on activities and aims in strategic competence development projects. The tentative model with its two perspectives is described and discussed in this article as a preliminary result. The model includes four identified key entities and their roles in pedagogical digital competence development; academic departments and their faculty, educational developers, infrastructure and IT-department and the pedagogical research unit. Further, a process model based on existing support structures, complemented with activities that can be sustained after the HEaD project ends is presented.

Wireless visual sensor networks provide featurerich information about their surrounding and can thus be used as a universal measurement tool for a great number of applications. Existing solutions, however, have mainly been focused on high sample rate applications, such as video surveillance, object detection and tracking. In this paper, we present a wireless camera node architecture that targets low sample rate applications (e.g., manual inspections and meter reading). The major design considerations are a long system lifetime, a small size and a low production cost.We present the overall architecture with its individual design choices, and evaluate the architecture with respect to its application constraints. With a typical image acquisition cost of 1.5 J for medium quality images and a quiescent power demand of only 7 uW, the evaluation results demonstrate that long operation periods of the order of years can be achieved in low sample rate scenarios.

Wireless Sensor Networks have the ability to improve a multitude of existing application domains. These networks are built up from a number of sensor nodes with sensing, communication and processing capabilities and the performance of the networked system is defined by the performance of the node platform it is based on. In this paper, we present SENTIO-em, a hardware platform for research in the environmental monitoring application domain. Based on the application domain requirements, the architecture and implementation of SENTIO-em is optimized for environmental monitoring constraints, while it is sufficiently flexible to be reused for different applications within the domain. The architecture of the platform is presented and evaluated under both laboratory and different environmental conditions. The obtained results are compared to a number of existing node platforms, demonstrating that SENTIO-em provides high energy efficiency with increased processing performance, short state transition times, and low quiescent currents.

Indoor photovoltaic (PV) application gains in attraction for low-power electronic systems, which requires accurate methods for performance predictions in indoor environments. Despite this, the knowledge on the performance of commonly used photovoltaic device models and their parameter estimation techniques in these scenarios is very limited. Accurate models are an essential tool for conducting feasibility analyses and component dimensioning for indoor photovoltaic systems. In this paper, we therefore conduct a comparison of the one- and two-diode models with parameters estimated based on two well-known methods. We evaluate the models' performance on datasets of photovoltaic panels intended for indoor use, and illumination conditions to be expected in indoor environments lit by artificial light sources. The results demonstrate that the one-diode model outperforms the two-diode model with respect to the estimation of the overall I-V characteristics. The two-diode model results instead in lower maximum power point errors. Both models show a sensitivity to initial conditions, such as the selection of the diode ideality factor, as well as the curve form of the photovoltaic panel to be modeled, which has not been acknowledged in previous research.

Ambient light measurements and an understanding of light conditions are essential for the accurate estimation of available energy in indoor photovoltaic applications. Light conditions may vary with respect to illumination intensity, duration, and spectral composition. Although the importance of the light spectrum has been documented in laboratory studies, previous distributed measurement methods are limited to intensity as a measure for output power. In this paper, we propose and implement a system for distributed measurement of light conditions that includes spectral information with low overhead. Based on a prototype implementation, we demonstrate that the illumination intensity and spectrum varies considerably over time and space, which confirms the demand for the proposed solution. We, moreover, characterize the energy consumption of the prototype, demonstrating that long-term, unattended characterization of light conditions can be achieved. 

Solar energy harvesting allows for wireless sensor networks to be operated over extended periods of time. In order to select an appropriate harvesting architecture and dimension for its components, an effective method for the comparison of system implementations is required. System simulations have the capability to accomplish this in an accurate and efficient manner. In this paper, we evaluate the existing work on solar energy harvesting architectures and common methods for their modeling. An analysis of the existing approaches demonstrates a mismatch between the requirement of the task to be both accurate and efficient and the proposed modeling methods, which are either accurate or efficient. As a result, we propose a data-driven modeling method based on artificial neural networks for further evaluation by the research community. Preliminary results of an initial investigation demonstrate the capability of this method to accurately capture the behavior of a solar energy harvesting architecture, while providing a time-efficient model generation procedure based on system-level data.

Models of photovoltaic devices are used to compare the properties of photovoltaic cells and panels, and to predict their I-V characteristics. To a large extent, modeling methods are based on the one-diode equivalent circuit. Although much research exists on the implementation and evaluation of these methods for typical outdoor conditions, their performance at indoor illumination levels is largely unknown. Consequently, this work performs a systematic study of methods for the parameter extraction of one-diode models under indoor conditions. We selected, reviewed and implemented commonly used methods, and compared their performance at different illumination levels. We have shown that most methods can achieve good accuracies with extracted parameters regardless of the illumination condition, but their accuracies vary significantly when the parameters are scaled to other conditions. We conclude that the physical interpretation of extracted parameters at low illumination is to a large extent questionable, which explains errors based on standard scaling approaches. 

Existing solar energy harvesting systems are typically evaluated with a single configuration. However, results on different harvester configurations are often desired in order to select the appropriate match to specific ambient conditions and application requirements. In this paper, we therefore present a concept for remotely reconfigurable solar energy harvesting testbeds, which allows for multiple harvester configurations to be evaluated with a single system deployment. We demonstrate that such a testbed can be implemented in an efficient manner by utilizing the benefits of wireless sensor networks, resulting in a scalable and flexible system with low power consumption. 

In wireless sensor networks, as energy limited systems, communication is a costly activity. For this reason duty cycling approaches are commonly used, because they can limit the overall power consumption of a sensor node tremendously by shutting down communication sub-circuits whenever they are not used. However, for efficient power reduction nodes have to know the exact times when they are supposed to communicate in the network. Synchronization can be used to accomplish this and comes with additional features such as the possibility of cooperative sampling at a given time. In this paper we propose a synchronization protocol that introduces low overhead due to broadcast master-node synchronization, while still accomplishing synchronization accuracies in the order of 100 μs. The protocol is intended for periodic data collection applications that are common tasks in environmental monitoring systems. Since changes in environmental conditions can have a large effect on the synchronization behavior, we further present a temperature compensation algorithm for the proposed synchronization protocol that allows stable usage of synchronization in a wide range of temperatures. Measurement results are taken from implementing the protocol on sensor node platforms and show the real world performance of the presented methods.

There is a significant potential for Wireless sensor networks to be used as a general distributed measurement and monitoring system. The integration of computation, communication and sensing enables smart sensors to be built that can be adapted to a plethora of application requirements and allow for automated data collection throughout the network. However, the potential end users of this systems are domain experts, who usually do not possess the technical expertise to program, and thus operate, wireless sensor nodes, which prohibits the technology from becoming off-the-shelf equipment. In this paper, we present a method which enables the complexity of programming sensor nodes to be concealed in order to allow domain experts to use wireless sensor networks in basic applications without the requirement of technical assistance. We propose to use a computer-based specification entry, which generates a configuration parameter set to adjust the sensor node's application behavior. The method has been implemented in a proof-of-concept system and evaluated with test subjects who possess limited programming skills. The results show that users without any prior programming knowledge, or experience with embedded systems, are capable of configuring a sensor node according to a given application scenario within minutes.

Typical wireless sensor network applications inthe domain of environmental monitoring require or profitfrom extended system lifetime. However, restrictions in sensornode resources, especially due to the usage of capacity limitedbatteries, forbid these desired lifetimes to be reached. Asopposed to batteries, energy harvesting from ambient energysources enables for near-perpetual supply of sensor nodes, asthe utilized energy source is inexhaustible. Nevertheless, thesupply from ambient energy sources is rate-limited, whereinthis supply-rate is mainly defined by the system deploymentlocation. On the other hand, the attached sensor node hasa consumption-rate, which has to be supplied to guaranteecontinuous node operation. In this paper, we address thematching of supply-rate and consumption-rate in solar energyharvesting systems at locations with limited insolation. Thefocus lies on the reduction of harvester energy overhead, whichin low-duty cycled system easily reaches similar or higherconsumption levels than the load it supplies. We suggest andpresent two harvester architectures [1], that have their maindesign consideration on simplicity. The individual modulesof the architectures are tested and verified in laboratorymeasurements and we evaluate the fully implemented systemsin an outdoor deployment. Based on the laboratory results,implementation choices for the architecture modules have beenmade. Whereas both harvesting architectures continuouslysupplied the attached load during the deployment period, wewere able to compare their behavior with each other andpresent individual advantages and drawbacks

Environmental monitoring applications demand wireless sensor networks to operate over a long period of time. Although energy consumption of these systems has been tremendously reduced, lifetime of sensor nodes is still limited by the capacity and lifetime of batteries used as energy sources. Energy harvesting, and in outdoor deployments particular, solar energy harvesting becomes a suitable way of powering wireless sensor nodes as their power consumption decreases. In this paper we address the feasibility of battery-less operation of wireless sensor nodes using solar energy harvesting at locations where the amount of solar radiation is severely limited and seasonal variations are large. We present two circuit architectures optimized for low energy leakage and evaluate their performance based on data gathered in a deployment during winter in Sundsvall, Sweden. We show that both architectures allow operation of sensor nodes even in the darkest period of the year. Furthermore comparisons between the two architecture designs are provided. © 2010 IEEE.

With the advent of low-cost and low-power computation, communication and sensor devices, novel instrumentation and measurement applications have been enabled, such as real-time industrial condition monitoring and fine-grained environmental monitoring. In these application scenarios, a lack of available infrastructures for communication and power supply is a common problem. In industrial applications, for example, the machine to be monitored and the monitoring system itself have significantly different technology lifespans, which requires that the monitoring system be retrofitted to machines that are already in use. In environmental monitoring, measurement systems are deployed as standalone devices in potentially remote areas. Consequently, the more autonomous the sensor system can be in terms of required infrastructure, the better it can match application and business needs.

Solar energy harvesting has become a common energy source for outdoor wireless sensor networks. To avoid the lifetime limitation of traditional secondary battery technologies in these systems, energy harvesting architectures with short-term energy storage can be chosen. These technologies offer long shelf-life and many recharge cycles, but can buffer for only short periods of time due to their small storage capacity. In this paper we present the analysis of two of these short-term energy storage devices, namely double layer capacitors and thin-film batteries. We present different harvesting architectures using these buffer elements and compare their advantages and disadvantages in relation to being used in low-power wireless sensor network applications. Experimental results show that both storage types are viable options for the intended application, each bringing their own strengths and weaknesses.

In applications where a priori determination of location is infeasible, node localization schemes are desirable, which allow the node to estimate its location during network operation. The majority of these schemes are based on ranging between node pairs, which should ideally be performed without adding cost or size to the sensor node. Two-way time-of-flight schemes can fulfill this desire, by utilizing the measurement of the time-of-flight of electromagnetic waves to determine the distance between two sensor nodes. In this paper, we present the implementation and analysis of such a ranging scheme. Because a small error in time measurement can result in a large distance estimation error, the focus of this work lies on the determination and analysis of influencing factors, which limit the accuracy of round-trip-time measurements. We analyze two main contributing factors to the accuracy of the ranging scheme, namely the radio transceiver clock quantization and the link quality during round-trip-time measurement. These effects and their impact on the overall ranging error have been investigated by means of simulation and experimentation. Initial ranging errors as large as 24 m RMS were observed, which could be reduced to errors between 5 and 8 m RMS by utilizing compensation techniques.

Solar energy harvesting gains more and more attention in the field of wireless sensor networks. In situations, where these sensor systems are deployed outdoors, powering sensor nodes by solar energy becomes a suitable alternative to the traditional way of battery power supplies. Since solar energy, opposed to batteries, can be considered as an inexhaustible energy source, scavenging this source allows longer system lifetimes and brings wireless sensor networks closer to be an autonomous system with perpetual lifetime. Despite the possibility of designing and constructing these harvesting system, dimensioning becomes a crucial task to fit implemented components to application and load system demands. In this paper we present a way of dimensioning solar harvesting systems based on simulation. Method and implementation of component and system models are described on the basis of an example architecture that has been used in prior work. Furthermore we evaluate the model in comparison to deployment of the same architecture and show the suitability of using the simulation as a support to optimize choices for system parameters

In order to integrate equipment from different vendors, wireless sensor networks need to become more standardized. Using IP as the basis of low power radio networks, together with application layer standards designed for this purpose is one way forward. This research focuses on implementing and deploying a system using Contiki, 6LoWPAN over an 868 MHz radio network, together with CoAP as a standard application layer protocol. A system was deployed in the Cairngorm mountains in Scotland as an environmental sensor network, measuring streams, temperature profiles in peat and periglacial features. It was found that RPL provided an effective routing algorithm, and that the use of UDP packets with CoAP proved to be an energy efficient application layer. This combination of technologies can be very effective in large area sensor networks.

The behavior prediction of solar energy harvesting systems requires accurate system models in order to dimension the system with respect to its application and location constraints. In contrast to commonly used equivalent circuit models, artificial neural networks (ANN) allow for the behavior of the entire system to be captured in an efficient manner. In this work, we have investigated the influences of the underlying ANN configuration on the model's prediction performance. It was found that the performance variation between training runs increases with an rising number of neurons, which can lead to a higher model performance, but makes the performance outcome more sensitive to initial training conditions. Moreover, the results demonstrate that even simple ANN configurations capture the system behavior accurately and result in low prediction errors for the harvesting architecture under test.

Electromagnetic energy harvesting (EM) is an interesting method for obtaining energy from vibration sources to power autonomous wireless sensor systems. This is reached with the help of energy management circuits that can be based on voltage multipliers circuits or boost rectifier circuits. These energy management structures aim to raise the voltage of the EM converter to a level that is adequate for the storage element. In this paper, a dual-mode AC/DC circuit for EM is proposed for fast charging of the supercapacitor. The circuit is based on a boost rectifier and a voltage multiplier using discrete elements and without a startup circuit. This structure can work properly with an EM converter generating power in the range of a few µW. The circuit can work under different frequencies and accelerations. In the first step, the proposed circuit store the energy in a supercapacitor in voltage multiplier mode until it reaches 1.9 V. After that, the supercapacitor will supply the ancillary circuit to activate the boost mode based on bidirectional MOSFETs, which reduce the time to charge the supercapacitor. The maximum power consumption of the ancillary circuit is 34 µW. The experimental investigation shows that the system at acceleration ACC = 0.17 g and Freq = 27 Hz is able to charge the supercapacitor by a voltage of 5.41 V in 7 min. The proposed solution helps to enhance the time to charge comparing conventional voltage multiplier structure by 23 min.

Industrial condition monitoring has benefited significantly from developments in machine learning and deep learning. However, textile machines, to a large extent, still use simple sensor systems, requiring additional manual quality inspections. This paper focuses on applying deep neural networks (DNNs) in image-based condition monitoring of air-spinning machines. It specifically focuses on the spinning pressure parameter, which is strongly related to the quality of the produced yarn. The study aims to develop a method to detect structural defects in yarns and assign them to specific machine conditions. DNNs are used to analyze images of yarns generated at different spinning pressures within the spinning box to create a rich dataset for training deep learning models. The study then evaluates the effectiveness of the DNN-based approach in detecting and classifying structural defects in yarns and determining the corresponding machine conditions. The results demonstrate that the developed model can distinguish good yarn from bad yarn, which is used to analyze the proportion of good yarn segments in a longer yarn section. A decreasing proportion with decreasing spinning pressure can thus be used to identify trends in degrading machine conditions. The outcomes of the presented research could potentially help textile enterprises improve the quality and efficiency of their yarn manufacturing processes. 

The Internet of Things introduces Internet connectivity to things and objects in the physical world and thus enables them to communicate with other nodes via the Internet directly. This enables new applications that for example allow seamless process monitoring and control in industrial environments. One core requirement is that the nodes involved in the network have a long system lifetime, despite limited access to the power grid and potentially difficult propagation conditions. Energy harvesting can provide the required energy for this long lifetime if the node is able to send the data based on the available energy budget. In this paper, we therefore analyze and evaluate which common IoT communication technologies are suitable for nodes powered by energy harvesters. The comparison includes three different constraints from different energy sources and harvesting technologies besides several communication technologies. Besides identifying possible technologies in general, we evaluate the impact of duty-cycling and different data sizes. The results in this paper give a road map for combining energy harvesting technology with IoT communication technology to design industrial sensor nodes. 

In typical application scenarios of wireless sensor systems the embedded software follows a sequential program-flow. Such sequential systems are easily described using a finite state machine (FSM) abstraction model, as it is known from digital circuit design and control system development. In practice, a theoretical description in the form of an FSM must be transferred into a software implementation using an efficient programming abstraction method. This paper presents the implementation of a lightweight software library, which supports the development of wireless sensor applications based on hierarchical finite state machines (HFSM). With its minimalistic design approach, the implementation creates a small memory footprint and a high efficiency in state transitions and in Interrupt Service Routine execution. By introducing a hierarchical state machine architecture it is possible to split a complex algorithm into sub-modules. Hence, the code modularity and reusability is improved. Moreover, it is possible to test each module independently, so that code errors can be traced back more rapidly. This results in a very reliable application software.

For the output power estimation of photovoltaic devices in indoor applications, models are needed that perform accurately at the low illumination levels encountered. As a robust and scalable solution, we propose a data-driven modeling method, spanning an interpolated surface between two reference I-V curves. The proposed approach is evaluated based on experimental data of two exemplar PV panels at indoor illumination levels. The results are compared to two common parameter extraction methods for the one-diode circuit model. This investigation demonstrates that the proposed surface model has a high performance under all test conditions, whereas the reference models show a performance dependency on the PV panel type. It can be concluded that the surface model is a competitive alternative for output power estimations at indoor illumination levels, removing many of the uncertainties of traditionally used physical parameter extraction and scaling methods.

The characterization of light conditions plays an important role in the estimation of available energy levels to ambient light energy harvesting systems. Indoor light conditions are commonly described by illuminance levels. The same illuminance levels, however, can be generated by different light source types, which radiate different spectral components. This means that based on their spectral response, solar panels can produce different output powers even though identical illuminance levels are observed. We propose a method to distinguish these conditions based on limited spectral information. Using low-cost sensors, spectral characteristics of the light condition can be acquired and used to classify the underlying light source type, which allows for a more accurate estimation of the solar panel response. The method was evaluated experimentally for a number of common indoor light sources and under different conditions. Evaluation results have shown that the method can be used to distinguish the light sources under test with very high classification accuracy. Moreover, the method can be used accurately in situations with limited interference. This makes it a low-cost alternative to the characterization of light conditions using spectrometers, the use of which is infeasible in spatially distributed characterization applications.

Ambient light energy harvesting is a cost-effective and mature approach for supplying low-power sensor systems with power in many indoor applications. Although the spectral information of a light source is known to influence the efficiency and output power of a photovoltaic cell, the spectrum of the ambient illumination is due to measurement complexity often neglected when characterizing light conditions for power estimation purposes. In this paper we evaluate the influence of considering spectral information on the energy estimation accuracy. We create a dataset of varying light conditions in a typical indoor environment based on eight locations. For each location, we compare the energy estimation accuracy with and without spectral considerations. The results of this investigation demonstrate that a spectrum-based method leads to significant performance improvements in cases where the light condition is not defined by a single light source.

Models of photovoltaic devices are an important tool for the estimation of their I-V characteristics. These characteristics, in turn, can be used to optimize production, compare devices, or predict the output power under different illumination conditions. Equivalent circuit models are the most common model types utilized. Although these models and the estimation of their parameters are thoroughly investigated, little is known about their performance under indoor illumination conditions. This, however, is essential for applications where photovoltaic devices are used indoors, such as for PV-powered sensors, wearables or Internet of Things devices. In this paper, a comprehensive and quantitative study of parameter estimation methods for the two-diode model is conducted, focusing particularly on the performance at indoor illumination levels. We reviewed and implemented a set of six common parameter estimation methods, and evaluate the performance of the estimated parameters on a typical photovoltaic module utilized in indoor scenarios. The results of this investigation demonstrate that there is a large performance variation between different parameter estimation methods, and that many methods have difficulties to estimate accurate parameters at low illumination conditions. Moreover, the majority of methods result in physically infeasible parameters, at least under some of the evaluated conditions. When applying physically motivated parameter scaling methods to these parameters, large estimation errors are observed, which limits the model’s applicability for power estimation purposes. 

The growing interest in indoor light energy harvesting for wireless sensor systems and low-power electronics has created a demand for systematic design methods that optimize the system implementation and component choices for different lighting scenarios. Although the spectrum of light is known to influence the efficiency and output power of a photovoltaic cell, existing power estimation methods neglect the spectrum. By contrast, the power estimation method proposed in this paper takes spectral and intensity information into account. It uses low-cost sensors to measure spectral information and select an appropriate device model based on the classification of the light source. The method is evaluated under different light conditions, including individual light sources, mixed artificial light sources, and mixtures of artificial light and sunlight. The results demonstrate that the proposed implementation selects a reasonable model in most cases, including mixed light source conditions. Using light source specific models for photovoltaic panels, the resulting estimation error is low and has clear advantages over methods neglecting spectral information.

In the energy harvesting domain, the modelling of a solar panel plays an important role in predicting the energy availability of energy harvesting system applications. Indoor environments, which are illuminated by artificial light sources, have typically much lower illumination levels than outdoor environments. In this paper, we compare the behaviour of different types of models under low illuminance conditions, in order to investigate sufficient modelling approaches for indoor environments. Previous work has shown that equivalent circuit modelling may have reduced performance under low illuminance conditions. Instead, we investigate behavioural models and compare their results with the equivalent circuit model. Two different types of behavioural models have been tested, namely artificial neural network models and polynomial curve fitting models. The comparison of these three models has shown that it is not possible to establish which of the modelling methods performs best, because each of them have strong points and shortcomings making the ideal choice application dependant.

Precision livestock farming allows farmers to optimize herd management while significantly reducing labor needs. Individualized monitoring of cattle feeding behavior offers valuable data to assess animal performance and provides valuable insights into animal welfare. Current acoustic foraging activity recognizers achieve high recognition rates operating on computers. However, their implementations on portable embedded systems (for use on farms) need further investigation. This work presents two embedded deployments of a state-of-the-art foraging activity recognizer on a low-power ARM Cortex-M0+ microcontroller. The parameters of the algorithm were optimized to reduce power consumption. The embedded algorithm processes masticatory sounds in real-time and uses machine-learning techniques to identify grazing, rumination and other activities. The overall classification performance of the two embedded deployments achieves an 84% and 89% balanced accuracy with a mean power consumption of 1.8 mW and 12.7 mW, respectively. These results will allow this deployment to be integrated into a self-powered acoustic sensor with wireless communication to operate autonomously on cattle. 

Precision livestock farming (PLF) leverages cutting-edge technologies and data-driven solutions to enhance the efficiency of livestock production, its associated management, and its welfare. Continuous monitoring of the masticatory sound of cattle allows the estimation of dry-matter intake, classification of jaw movements (JMs), and recognition of grazing and rumination bouts. Over the past two decades, algorithms for analyzing feeding sounds have seen improvements in performance and computational requirements. Nevertheless, in some cases, these algorithms have been implemented on resource-constrained electronic devices, limiting their functionality to one specific task: either classifying JMs or recognizing feeding activities (such as grazing and rumination). In this work, we present an acoustic monitoring system that comprehensively analyzes grazing cattle's feeding behavior at multiple scales. This embedded system classifies different types of JMs, identifies feeding activities, and provides predictor variables for estimating dry-matter intake. Results are transmitted remotely to a base station using long-range communication (LoRa). Two variants of the system have been deployed on a Raspberry Pi Pico board, based on a low-power ARM Cortex-M0+ microcontroller. Both firmware versions make use of direct access memory, sleep mode, and clock-gating techniques to minimize energy consumption. In laboratory experiments, the first deployment consumes 20.1 mW and achieves an F1-score of 87.3% for the classification of JMs and 87.0% for feeding activities. The second deployment consumes 19.1 mW and reaches an F1-score of 84.1% for JMs and 83.5% for feeding activities. The modular design of the proposed embedded monitoring system facilitates integration with energy-harvesting power sources for autonomous operation in field conditions.

Electro-mechanical systems operating in periodic cycles are pivotal in the Industry 4.0, enabling automated processes that enhance efficiency and productivity. Early detection of failures and anomalies in duty cycles of these machines is crucial to ensure uninterrupted operation and prevent costly downtimes. Although the wear and damage of machines have been extensively studied, a significant proportion of these problems can be traced back to operator errors, underlining the importance of continuously monitoring the machine activity to ensure optimal performance. This work presents an automatic algorithm designed to identify improper duty cycles of industrial machines, exemplified on a mining conveyor belt. To enable the identification of duty cycles, the operational states of the machine are first categorized using machine learning (ML). The study compares six tiny ML techniques on two resource-constrained microcontrollers, reporting an f1-score of 87.6% for identifying normal and abnormal duty cycles and 96.8% for the internal states of the conveyor belt system. Deployed on both low-power microcontrollers, the algorithm processes input data in less than 106 μs, consuming less than 1.16 μJ. These findings promise to facilitate integration into more comprehensive preventive maintenance algorithms. 

As pointed out by many researchers, the ongoing pandemic has been a catalyst for educational development. With the increasing need for reskilling and lifelong learning, the current model of technology-enhanced learning needs updating, and so does also the university programmes for bachelor's and master's students. This study is based on an online brainstorming session and submitted development plans in the HEaD (Higher Education and Digitalisation) project, a five-year initiative for technology-enhanced educational development. HEaD is a development project aimed at supporting university teachers to work with research and development in the field of technology-enhanced and lifelong learning. As the research strategy, an action research approach was used, with the purpose of improving the educational process where authors also have the roles of teachers and facilitators. The aim of the study is to describe and discuss pilot project members' perceptions of challenges, opportunities, tools and techniques in higher education development. Data gathered from workshop discussion summaries and project plans were thematically analysed. Ideas from the workshop sessions were written down and saved with the use of the digital notice board Padlet. Results from the thematic analysis have been grouped into the four predefined categories of challenges, opportunities, tools and techniques. Findings show that course participants and project members have interesting ideas that have the potential to reinforce the current educational model at the university. Several tools and techniques that could support synchronous as well as asynchronous online learning will be tested and evaluated. Both the workshop summaries and the project plans show a high degree of creativity, but on the other hand, the method descriptions were scarce and would need improvement. The conclusion is that the project has had a good start if seen as development, but that there is a need for improvement and more input to achieve the intended core idea of research and development

The introduction of nonlinearities into energy harvesting in order to improve the performance of linear harvesters has attracted a lot of research attention recently. The potential benefits of nonlinear harvesters have been evaluated under sinusoidal or random excitation. In this paper, the performances of electromagnetic energy harvesters with linear and nonlinear springs are investigated under real vibration data. Compared to previous studies, the parameters of linear and nonlinear harvesters used in this paper are more realistic and fair for comparison since they are extracted from existing devices and restricted to similar sizes and configurations. The simulation results showed that the nonlinear harvester did not generate higher power levels than its linear counterpart regardless of the excitation category. Additionally, the effects of nonlinearities were only available under a high level of acceleration. The paper also points out some design concerns when harvesters are subjected to real vibrations. © 2020 by the authors. Licensee MDPI, Basel, Switzerland.

Investigating the coil–magnet structure plays a significant role in the design process of the electromagnetic energy harvester due to the effect on the harvester’s performance. In this paper, the performance of four different electromagnetic vibration energy harvesters with cylindrical shapes constrained in the same volume were under investigation. The utilized structures are (i) two opposite polarized magnets spaced by a mild steel; (ii) a Halbach array with three magnets and one coil; (iii) a Halbach array with five magnets and one coil; and (iv) a Halbach array with five magnets and three coils. We utilized a completely automatic optimization procedure with the help of an optimization algorithm implemented in Python, supported by simulations in ANSYS Maxwell and MATLAB Simulink to obtain the maximum output power for each configuration. The simulation results show that the Halbach array with three magnets and one coil is the best for configurations with the Halbach array. Additionally, among all configurations, the harvester with two opposing magnets provides the highest output power and volume power density, while the Halbach array with three magnets and one coil provides the highest mass power density. The paper also demonstrates limitations of using the electromagnetic coupling coefficient as a metric for harvester optimization, if the ultimate goal is maximization of output power. 

A new scheme for the automated design and optimization of electromagnetic energy harvesters is presented. The proposed method aims to deal with the limitations of current design techniques and to improve the efficiency of the design process. Most of the current design approaches require significant user experience and knowledge to make informed design decisions, which means that a large amount of time is needed to obtain the optimized design. Additionally, the design solution is suitable only for a specific application, which makes it difficult to adopt it to other application demands. In the proposed method, the development cycle has been sped up significantly by minimizing human efforts and utilizing a generic model for the design process. The method is based on a generic template based on specific resources and requirements, which is processed by a black-box optimization algorithm to come up with a number of promising configuration suggestions. The method's effectiveness and its autonomous operation are demonstrated based on the design optimization of an electromagnetic pick-up unit for vibration energy harvesters using a Halbach array.

Energy harvesting has been proposed to support or replace battery-based energy supplies in low-power loT systems. However, manual design and optimization is currently required to implement energy harvesters. This implies significant time and cost requirements, reducing the competitiveness of energy harvesting solutions in many applications. In this paper, we propose and study a method for the automated design of ortho-planar springs in vibration energy harvesters. The proposed method uses a Python tool that has been developed to translate parametric descriptions of the spring into 2D and 3D spring CAD models. The tool is combined with a black-box optimization algorithm in order to identify optimized spring parameters. An evaluation of the proposed method on a case study of an electromagnetic vibration energy harvester demonstrates that the approach successfully results in spring designs that perform well in the energy harvesting application. Consequently, the requirement of expert competence to adjust an energy harvester to new application constraints is significantly reduced, contributing to the availability and competitiveness of energy harvesting solutions in real world applications.