Electrical components constantly being scaled down in size allows for small, inexpensive sensors to be placed on or around the human body for motion sensing applications. In addition, the merging of textiles with electrical components, known as electronic textiles (e-textiles), allows for these sensors to be placed directly on a wearable fabric. Simulation allows for extensive application testing and verification before prototype development. This thesis presents a simulation environment for motion sensing E-textiles. Specifically, this environment incorporates motion capture position data to simulate a rate sensing gyroscope and a dual-axis accelerometer. In addition, this simulation environment is applied to the field of gait analysis, which is the process of quantification and interpretation of a person's stride, to calculate a subject's step length. / Master of Science

ITC/USA 2008 Conference Proceedings / The Forty-Fourth Annual International Telemetering Conference and Technical Exhibition / October 27-30, 2008 / Town and Country Resort & Convention Center, San Diego, California / The US Army Research Laboratory has been involved in the design and implementation of electromagnetic gun technology for the past several years. One of the primary factors of this research is an accurate assessment of in-bore structural loads on the launch projectiles. This assessment is essential for the design of mass-efficient launch packages for electromagnetic guns. If not properly accounted for, projectile failure can result. In order to better understand the magnitude of the in-bore loads, a data-recorder was integrated with an armature and on-board payload that included tri-directional accelerometers and magnetic field sensors. Several packages were launched from an electromagnetic railgun located at Aberdeen Proving Ground, MD. Substantial effort was placed on soft-catching the rounds in order to facilitate data recovery. Analysis of the recovered data provided acceleration and magnetic field data acquired during the launch event.

Pesquisas em Interfaces Naturais, sub-área da Computação Ubíqua, investigam o uso de dispositivos não-tradicionais para possibilitar a interação entre usuários e aplicaçõs de maneiras menos intrusivas (gestos, voz e escrita baseada em tinta eletrônica, por exemplo). Com o aumento da popularidade de dispositivos equipados com sensores de aceleração, os desenvolvedores agora dispõem de um novo dispositivo que pode ser utilizado para prover interação entre usuários e diferentes aplicações, como por exemplo as que se encontram presentes em ambientes de TV interativos. Assim, aplicações que fazem uso de acelerômetros vêm sendo desenvolvidas para situações específicas, e suas implementações e formatos de dados manipulados são dependentes do domínio para o qual foram projetados. Este trabalho apresenta um modelo para a formalização do modo como esses dados podem ser manipulados, por meio de uma abordagem genérica e extensível. Além disso, o modelo permite a descrição de regras para agregação de valor a estes dados por meio da adição de significados. Isto e obtido com a proposta de uma arquitetura em camadas que possibilita a estruturação e compartilhamento desses dados de modo flexível. Três protótipos foram implementados na linguagem de programação Java, fazendo-se uso dessa arquitetura e de uma API desenvolvida para facilitar o uso do modelo. Essas implementações demonstram a viabilidade do modelo proposto como solução para a questão da interoperabilidade nos cenários ilustrados, e para a extensibilidade dos dados, nos casos em que uma mudança de requisitos faz-se necessária / Research in Natural Interfaces, sub-area of Ubiquitous Computing, investigates the use of non-traditional devices to support user interaction with applications in less intrusive ways (gestures, voice and writing based on electronic ink, for instance). With the increasing popularity of accelerometers, developers now have another tool that can be used to provide interaction between users and different applications, such as interactive TV environments. However, applications that make use of accelerometers are currently being developed for specific situations, and their implementations and handled documents are also dependent on the domain for which they were designed. This work aims to propose a model to formalize how the accelerometer data may be handled in a generic way. In addition, the model enables the description of rules to aggregate value to these data through the addition of meanings. This is done by proposing a layered architecture to structure and share data in a exible way. Three prototypes were implemented in the Java programming language, making use of this architecture and an API designed to facilitate the model implementation. These prototypes demonstrate the feasibility of the model proposed as a solution to the issue of interoperability in the scenarios illustrated, and the problem of data extensibility, whenever a change of requirements poses necessary

Active control constitutes the state of the art in vibration management in rotating machines. However, existing designs are impractical and costly, and hence not yet widely applied. The goal of the research reported here was to develop a design which would allow the implementation of active technology in a wider range of rotating machine applications. Thus, this study focuses on a novel active rotor topology, consisting of a hollow rotor with internally mounted sensors and actuators. This layout provides greater freedom to select the sensor and actuator positions along the rotor, and naturally protects the devices from harsh working environments. The research was structured according to four themes. Firstly, the concept feasibility was explored by constructing a fully functioning prototype. MEMS accelerometers and mass balancer actuators were mounted in an assembled rotor, together with a microcontroller and radio unit to enable wireless transmission of data. Secondly, the behaviour of MEMS accelerometers in a rotating frame of reference was studied. An output model was derived and applied to the study of whirl orbits and transient vibration. Further, techniques were developed to extract mean displacement and angular velocity information from the sensor signals. An analysis of potential sources of measurement error was conducted, and methods for their mitigation devised. The third theme focused on developing active vibration control techniques suitable for use with active rotors. The core of this work is the development and successful implementation of a non a priori method, Algorithmic Direct Search Control. This technique enables vibration to be minimised without knowledge of the system characteristics, by applying a direct search optimisation technique as a control law. Finally, the combination of active rotors and Active Magnetic Bearings was considered to tackle the problem of sensor/actuator non-collocation. The challenge of levitating a rotor on AMBs using only internal accelerometers was approached via integration-based displacement information extraction, to exploit existing PID controllers. This method proved unfeasible in practice, but valuable lessons were derived from the study. The key finding of this work is that active rotor technology is versatile, cost-effective, powerful and feasible. As such, it offers great potential as a route to achieving a more practical and generalised implementation of active control technology in rotating machinery.

The use of wearable sensors in consumer health and medicine is a rapidly developing topic of interest. The main purpose of the series of studies in this thesis is to identify novel uses of technology that can provide clinicians and scientists clinically feasible, low cost approaches to obtain meaningful information about functional limb symmetry in patients with knee injuries. In Study 1, individuals undergoing knee surgery were evaluated as they walked and stepped down onto a force platform in a manner similar to how one would step off a curb to cross a street. When subjects stepped onto their uninvolved leg, peak vertical ground reaction force was greater and occurred earlier than when stepping onto their involved leg. Asymmetries were greater in those with higher quadriceps neuromuscular impairment. In Study 2, the reliability and validity of using wearable accelerometer sensors was evaluated for estimating single leg vertical hop height in healthy people and individuals after ACL reconstruction surgery. The reliability and concurrent validity of using accelerometers to estimate single leg hop height were excellent, and were similar for healthy and ACL-reconstructed subjects. Error for this method was low, in particular when the accelerometer was worn at the lower leg. Asymmetry in hop height was greater in those with higher quadriceps neuromuscular impairment. In Study 3, wearable accelerometers were compared to a system of motion capture cameras and force platform as a method to assess functional movement asymmetry in healthy people and individuals after ACL reconstruction. While walking and stepping down, accelerometers worn at the waist were able to detect underlying movement asymmetry when it exists in people after ACL reconstruction. Acceleration at the waist was strongly associated with vertical ground reaction force and moderately associated with knee extension moments. Collectively, these studies provide evidence that functional movement symmetry can be measured with simple, inexpensive methods that can be used in a variety of clinical or field-based settings.

The study of physical activity is important in improving people���s health as it can help people understand the relationship between physical activity and health. Accelerometers, due to its small size, low cost, convenience and its ability to provide objective information about the frequency, intensity, and duration of physical activity, has become the method of choice in measuring physical activity. Machine learning algorithms based on the featurized representation of accelerometer data have become the most widely used approaches in physical activity prediction. To improve the classification accuracy, this thesis first explored the impact of the choice of data (raw vs processed) as well as the choice of features on the performance of various classifiers. The empirical results showed that the machine learning algorithms with strong regularization capabilities always performed better if provided with the most comprehensive feature set extracted from raw accelerometer signal. Based on the hypothesis that for some time series, the most discriminative information could be found at subwindows of various sizes, the Subwindow Ensemble Model (SWEM) was proposed. The SWEM was designed for the accelerometer-based physical activity data, and classified the time series based on the features extracted from subwindows. It was evaluated on six time series datasets. Three of them were accelerometer-based physical activity data, which the SWEM was designed for, and the rest were different types of time series data chosen from other domains. The empirical results indicated a strong advantage of the SWEM over baseline models on the accelerometerbased physical activity data. Further analysis confirmed the hypothesis that the most discriminative features could be extracted from subwindows of different sizes, and they were effectively used by the SWEM. / Graduation date: 2013

This dissertation was designed to determine the effects of body mass index (BMI) and walking speed on activity monitor outputs. A secondary purpose was to compare the activity monitors’ performance in a free-living environment. In the first experiment, 71 participants wore three waist-mounted activity monitors (Actical, ActiGraph, and NL-2000) and an ankle-mounted device (StepWatch 3) while walking on a treadmill (40, 67 and 94 m/min). The tilt angle of each device was measured. The Actical recorded 26% higher activity counts (P < 0.01) in obese persons with a tilt <10 degrees, compared to normal weight persons. The ActiGraph was unaffected by BMI or tilt angle. In the second experiment, the steps recorded by the devices were compared to actual steps. Speed had the greatest influence on the accuracy these devices. At 40 m/min, the ActiGraph was the least accurate device for normal weight (38%), overweight (46%) and obese (48%) individuals. The Actical, NL-2000 and StepWatch averaged 65%, 73% and 99% of steps taken, respectively. Lastly, several generations of the ActiGraph (7164, GT1M, and GT3X), and other research grade activity monitors (Actical; ActivPAL; and Digi-Walker) were compared to a criterion measure of steps. Fifty-six participants performed treadmill walking (40, 54, 67, 80 and 94 m/min) and wore the devices for 24-hours under free-living conditions. BMI did not affect step count accuracy during treadmill walking. The StepWatch, PAL, and the AG7164 were the most accurate across all speeds; the other devices were only accurate at the faster speeds. In the free-living environment, all devices recorded about 75% of StepWatch-determined steps, except the AG7164 (99%). Based on these findings, we conclude that BMI does not affect the output of these activity monitors. However, waist-borne activity monitors are highly susceptible to under-counting steps at walking speeds below 67 m/min, or stepping rates below 100 steps/min. An activity monitor worn on the ankle is less susceptible to these speed effects and provides the greatest accuracy for step counting.

The relationship between students’ physical activity (PA) and on-task behavior in the classroom setting was examined. Given that students spend nearly half of their waking hours in school, researchers have suggested that the school environment could play a crucial role in increasing children’s PA. Physical activity of 157 first- and second-grade students was assessed using ActiGraph (GT1M) accelerometers during school hours. Momentary time sampling (MTS) tracked the on-task behaviors of 72 of the 157 participants every 30 seconds. Multiple linear regressions and paired sample t tests were run to measure students’ classroom PA steps, PA intensity levels in the classroom, and on-task behaviors. Results indicated weak, yet significant, inverse correlations between students’ PA steps, PA intensity levels, and on-task behaviors (R =.40, R2 = .16, p = .01). On-task behaviors and steps taken in the classroom before recess indicate a significant inverse relationship (R = -.18, R2 = .03), indicating the pre-recess classroom steps account for 3% of the variance in on-task behavior. Steps taken in the classroom indicate a significant inverse relationship (R = -.20, R2 = .04) with on-task behavior. The results from the linear regression analysis after recess indicate that the post-recess steps can account for approximately 4% of the variance of the on-task behavior. These overall results suggest that greater PA levels in the classroom setting were associated with less on-task behavior. Results from the t test indicate a significant (t(143) = -4.32, p < .001) increase in on-task behavior (3%) after recess. In conclusion, the present study demonstrated that 84% of the variance in on-task behavior is accounted for by non-PA suggesting that other variables may affect students’ on-task behaviors in the classroom setting.

The assessment of physical activity is of great interest to the public and to researchers. One of the most popular and objective methods of assessing physical activity is using motion sensors such as pedometers and accelerometers. The purpose of this study was to assess the accuracy and reliability of step counting and energy expenditure estimation of two popular electronic pedometers and one accelerometer during running. In all step-counting trials, actual step counts were determined using a hand tally counter. Twenty subjects (18-40 years) were asked to volunteer to participate in the study. During the initial assessment subjects ran at five, six, seven, and eight miles per hour on a motor driven treadmill for ten minutes at each speed. While running they wore two pedometers and an accelerometer while oxygen consumption was measured. In the second assessment, all subjects ran a timed, self-paced mile around a 200-meter track while wearing all three activity monitors. In the third assessment, thirteen subjects ran at five, six, seven and eight miles per hour on a motor driven treadmill for three minutes at each speed to assess the step counting reliability of the devices. The three devices were accurate to within 1% at all speeds on the treadmill and on the self-paced track run. In addition, all three devices were reliable (r = 0.80). Caloric expenditure estimation accuracy the NL-2000 was ± 21% when expressed in both net and gross kcals. The Digi-walker was f 22% when assumed to be net kcals and ± 45% when assumed to be gross kcals. The Actigraph was accurate to ± 49% in terms of net calories. The results of the current study have shown that these devices are accurate and reliable for step counting at speeds between five and eight mph. However, their ability to assess caloric expenditure is not accurate at the same speeds. In conclusion, the application of these devices to runners is limited to highly accurate and reliable step counters. / School of Physical Education, Sport, and Exercise Science

Thesis (Ph. D.)--University of Oregon, 2008. / Typescript. Includes vita and abstract. "This dissertation includes unpublished co-authored material"--P. v. Includes bibliographical references (leaves 119-128). Also available online in Scholars' Bank; and in ProQuest, free to University of Oregon users.