This paper presents implementation details, system characterization, and the performance of a wearable sensor network that was designed for human activity analysis. Specific machine learning mechanisms are implemented for recognizing a target set of activities with both out-of-body and on-body processing arrangements. Impacts of energy consumption by the on-body sensors are analyzed in terms of activity detection accuracy for out-of-body processing. Impacts of limited processing abilities for the on-body scenario are also characterized in terms of detection accuracy, by varying the background processing load in the sensor units. Impacts of varying number of sensors in terms of activity classification accuracy are also evaluated. Through a rigorous systems study, it is shown that an efficient human activity analytics system can be designed and operated even under energy and processing constraints of tiny on-body wearable sensors.
-This paper presents the key concepts, system architecture, implementation details, and performance of an accelerometer-equipped bottle for monitoring and tracking liquid intake. The key system component is an elastic band, equipped with sensor and other electronics, which can be attached to a regular water bottle in order to track the bottle's usage movements. The software running on the band captures and detects acceleration signatures that the bottle experiences specifically during drinking events. Detecting such drinking events can lead to higher level monitoring such as tracking the consumed liquid volume. A Bluetooth based wireless link out of the electronic band is used for sending the detected drinking events to a smartphone or to a notebook computer for higher level tracking and data management. Different machine learning methods were adopted and experimented with for both drinking event detection and intake volume estimation. Through experiments on nine healthy subjects, the system is shown to be able to achieve up to 99% accuracy in drinking event detection, and up to 75% accuracy for intake volume estimation.
Increased mobility of hens in noncaged housing presents possibilities for bone breakage due to crash landings from jumps or flights between perches or housing infrastructure. Because bone breakage is a welfare and economic concern, understanding how movement from different heights affects hen landing impact is important. By tracking 3-dimensional bird movement, an automated sensor technology could facilitate understanding regarding the interaction between noncage laying hens and their housing. A method for detecting jumps and flight trajectories could help explain how jumps from different heights affect hen landing impact. In this study, a wearable sensor-based jump detection mechanism for egg-laying hens was designed and implemented. Hens were fitted with a lightweight (10 g) wireless body-mounted sensor to remotely sample accelerometer data. Postprocessed data could detect occurrence of jumps from a perch to the ground, time of jump initiation, time of landing, and force of landing. Additionally, the developed technology could estimate the approximate height of the jump. Hens jumping from heights of 41 and 61 cm were found to land with an average force of 81.0 ± 2.7 N and 106.9 ± 2.6 N, respectively, assuming zero initial velocity (P < 0.001). This paper establishes the technological feasibility of using body-mounted sensor technology for jump detection by hens in different noncage housing configurations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.