This paper reports on the first stage of a research project1) that aims to incorporate objective measures of physical activity into health and lifestyle surveys. Physical activity is typically measured with questionnaires that are known to have measurement issues, and specifically, overestimate the amount of physical activity of the population. In a lab setting, 40 participants wore four different sensors on five different body parts, while performing various activities (sitting, standing, stepping with two intensities, bicycling with two intensities, walking stairs and jumping). During the first four activities, energy expenditure was measured by monitoring heart rate and the gas volume of in‐ and expired O2 and CO2. Participants subsequently wore two sensor systems (the ActivPAL on the thigh and the UKK on the waist) for a week. They also kept a diary keeping track of their physical activities, work and travel hours. Machine learning algorithms were trained with different methods to determine which sensor and which method was best able to differentiate the various activities and the intensity with which they were performed. It was found that the ActivPAL had the highest overall accuracy, possibly because the data generated on the upper tigh seems to be best distinguishing between different types of activities and therefore led to the highest accuracy. Accuracy could be slightly increased by including measures of heartrate. For recognizing intensity, three different measures were compared: allocation of MET values to activities (used by ActivPAL), median absolute deviation, and heart rate. It turns out that each method has merits and disadvantages, but median absolute deviation seems to be the most promishing metric. The search for the best method of gauging intensity is still ongoing. Subsequently, the algorithms developed for the lab data were used to determine physical activity in the week people wore the devices during their everyday activities. It quickly turned out that the models are far from ready to be used on free living data. Two approaches are suggested to remedy this: additional research with meticulously labelled free living data, e.g., by combining a Time Use Survey with accelerometer measurements. The second is to focus on better determining intensity of movement, e.g., with the help of unsupervised pattern recognition techniques. Accuracy was but one of the requirements for choosing a sensor system for subsequent research and ultimate implementation of sensor measurement in health surveys. Sensor position on the body, wearability, costs, usability, flexibility of analysis, response, and adherence to protocol equally determine the choice for a sensor. Also from these additional points of view, the activPAL is our sensor of choice.
DOCUMENT
This work provides a feasibility study on estimating the 3-D locations of several thousand miniaturized free-floating sensor platforms. The localization is performed on basis of sparse ultrasound range measurements between sensor platforms and without the use of beacons. We show that this task can be viewed as a specific type of pose graph optimization. The main challenge is robustly estimating an initial pose graph, that models the locations of sensor platforms. For this, we introduce a novel graph growing strategy that uses random sample consensus in alternation with non-linear refinement. The theoretical properties of our sensor cloud localization method are analyzed and its robustness is investigated using simulations. These simulations are based on inlier-outlier measurement models and focus on the application of subterranean 3-D mapping of liquid environments, such as pipe infrastructures and oil wells.
DOCUMENT
Improving estrus detection accuracy could improve sow conception rates,leading to higher production efficiency. Current observation-based estrusdetection practices are labor intensive and less accurate. Around estrus, bodytemperature and activity change. Therefore in this study a telemetric monitoringsystem for body temperature and activity was tested. Firstly Templant2 sensors(TeleMetronics) were validated under lab conditions for temperatures from 35°Cto 45°C, using a water basin with a Julabo heater and a P600 thermometer.Activity measurements were validated with the sensors attached to a stick,simulating sow movements. Secondly, sensors were attached externally to 4gilts and 4 sows for 30 minutes, testing functionality. Thirdly, activity of sowswas recorded manually for 3 days around estrus. Results showed that under labconditions temperature results of sensors, heater and thermometer were highlycorrelated (linear regression, R2=0,96; slope 1,1). Simulated activitiescorresponded consistently with peaks in sensor values. Activity was measuredreliably with the sensor attached externally to the sows. On the farm, sowsshowed more activity (manual observations, P<0.05 for standing up, lying down,sitting down and walking) the day before insemination. We conclude thatmonitoring activity and body temperature is a promising tool for estrousdetection in sows.
LINK
A low-cost sensornode is introduced to monitor the 5G EMF exposure in the Netherlands for the four FR1 frequency bands. The sensornode is validated with in-lab measurements both with CW signals as for QAM signals and perform for both cases and for all frequency bands an error less than 1 dB for a dynamic range of 40 dB. This sensor is a follow up of the earlier version of our previously developed sensor and have substantial improvements in terms of linearity, error, and stability.
DOCUMENT
To better control the growing process of horticulture plants greenhouse growers need an automated way to efficiently and effectively find where diseases are spreading.The HiPerGreen project has done research in using an autonomous quadcopter for this scouting. In order for the quadcopter to be able to scout autonomously accurate location data is needed. Several different methods of obtaining location data have been investigated in prior research. In this research a relative sensor based on optical flow is looked into as a method of stabilizing an absolute measurement based on trilateration. For the optical flow sensor a novel block matching algorithm was developed. Simulated testing showed that Kalman Filter based sensor fusion of both measurements worked to reduce the standard deviation of the absolute measurement from 30 cm to less than 1 cm, while drift due to dead-reckoning was reduced to a maximum of 11 cm from over 36 cm.
DOCUMENT
This work is on 3-D localization of sensor motes in massive swarms based solely on 1-D relative distance-measurements between neighbouring motes. We target applications in remote and difficult-to-access environments such as the exploration and mapping of the interior of oil reservoirs where hundreds or thousands of motes are used. These applications bring forward the need to use highly miniaturized sensor motes of less than 1 centimeter, thereby significantly limiting measurement and processing capabilities. These constraints, in combination with additional limitations posed by the environments, impede the communication of unique hardware identifiers, as well as communication with external, fixed beacons.
DOCUMENT
Knowing firefighters’ locations in a burning building would dramatically improve their safety. In this study from the Saxion Research Centre for Design and Technology in the Firebee project, an algorithm was developed and tested to enhance the estimation of a person’s location, based on inertial measurements combined with measurements of the earth’s magnetic field. The developed algorithm is an extension of the zero velocity update technique. Without any enhancements, the accuracy of the estimation is in the order of several meters after measuring for only a few seconds. With enhancements, the accuracy improved to be within five meters after measuring for ten minutes. Our result demonstrated that it is possible to determine in which room and on which floor a person is after ten minutes. Major improvements were observed in the estimation of the sensor’s height. The results are promising and the following phases of the project focus on improving the solution and on developing the concept into a practically applicable system.
MULTIFILE
This review offers a detailed examination of the current landscape of radio frequency (RF) electromagnetic field (EMF) assessment tools, ranging from spectrum analyzers and broadband field meters to area monitors and custom-built devices. The discussion encompasses both standardized and non-standardized measurement protocols, shedding light on the various methods employed in this domain. Furthermore, the review highlights the prevalent use of mobile apps for characterizing 5G NR radio network data. A growing need for low-cost measurement devices is observed, commonly referred to as “sensors” or “sensor nodes”, that are capable of enduring diverse environmental conditions. These sensors play a crucial role in both microenvironmental surveys and individual exposures, enabling stationary, mobile, and personal exposure assessments based on body-worn sensors, across wider geographical areas. This review revealed a notable need for cost-effective and long-lasting sensors, whether for individual exposure assessments, mobile (vehicle-integrated) measurements, or incorporation into distributed sensor networks. However, there is a lack of comprehensive information on existing custom-developed RF-EMF measurement tools, especially in terms of measuring uncertainty. Additionally, there is a need for real-time, fast-sampling solutions to understand the highly irregular temporal variations EMF distribution in next-generation networks. Given the diversity of tools and methods, a comprehensive comparison is crucial to determine the necessary statistical tools for aggregating the available measurement data.
MULTIFILE
Within rehabilitation, there is a great need for a simple method to monitor wheelchair use, especially whether it is active or passive. For this purpose, an existing measurement technique was extended with a method for detecting self- or attendant-pushed wheelchair propulsion. The aim of this study was to validate this new detection method by comparison with manual annotation of wheelchair use. Twenty-four amputation and stroke patients completed a semi-structured course of active and passive wheelchair use. Based on a machine learning approach, a method was developed that detected the type of movement. The machine learning method was trained based on the data of a single-wheel sensor as well as a setup using an additional sensor on the frame. The method showed high accuracy (F1 = 0.886, frame and wheel sensor) even if only a single wheel sensor was used (F1 = 0.827). The developed and validated measurement method is ideally suited to easily determine wheelchair use and the corresponding activity level of patients in rehabilitation.
DOCUMENT
