Locomotion analysis in horses and cows based on sensor signals

Deep learning models for human activity recognition (HAR) based on sensor data have been heavily studied recently. However, the generalization ability of deep models on complex real-world HAR data is limited by the availability of high-quality labeled activity data, which are hard to obtain. In this article, we design a similarity embedding neural network that maps input sensor signals onto real vectors through carefully designed convolutional and Long Short-Term Memory (LSTM) layers. The embedding network is trained with a pairwise similarity loss, encouraging the clustering of samples from the same class in the embedded real space, and can be effectively trained on a small dataset and even on a noisy dataset with mislabeled samples. Based on the learned embeddings, we further propose both nonparametric and parametric approaches for activity recognition. Extensive evaluation based on two public datasets has shown that the proposed similarity embedding network significantly outperforms state-of-the-art deep models on HAR classification tasks, is robust to mislabeled samples in the training set, and can also be used to effectively denoise a noisy dataset.

Download Full-text

Aerial Mapping of Odorous Gases in a Wastewater Treatment Plant Using a Small Drone

Remote Sensing ◽

10.3390/rs13091757 ◽

2021 ◽

Vol 13 (9) ◽

pp. 1757

Author(s):

Javier Burgués ◽

María Deseada Esclapez ◽

Silvia Doñate ◽

Laura Pastor ◽

Santiago Marco

Keyword(s):

Wastewater Treatment ◽

Wastewater Treatment Plants ◽

Treatment Plant ◽

Principal Component ◽

Hazardous Air Pollutants ◽

Sampling System ◽

Aerial Mapping ◽

Sensor Signals ◽

Current Monitoring ◽

Flexible Tubes

Wastewater treatment plants (WWTPs) are sources of greenhouse gases, hazardous air pollutants and offensive odors. These emissions can have negative repercussions in and around the plant, degrading the quality of life of surrounding neighborhoods, damaging the environment, and reducing employee’s overall job satisfaction. Current monitoring methodologies based on fixed gas detectors and sporadic olfactometric measurements (human panels) do not allow for an accurate spatial representation of such emissions. In this paper we use a small drone equipped with an array of electrochemical and metal oxide (MOX) sensors for mapping odorous gases in a mid-sized WWTP. An innovative sampling system based on two (10 m long) flexible tubes hanging from the drone allowed near-source sampling from a safe distance with negligible influence from the downwash of the drone’s propellers. The proposed platform is very convenient for monitoring hard-to-reach emission sources, such as the plant’s deodorization chimney, which turned out to be responsible for the strongest odor emissions. The geo-localized measurements visualized in the form of a two-dimensional (2D) gas concentration map revealed the main emission hotspots where abatement solutions were needed. A principal component analysis (PCA) of the multivariate sensor signals suggests that the proposed system can also be used to trace which emission source is responsible for a certain measurement.

Download Full-text

A Novel Method for Harmonic Current Injection with Sensor Supported Adaptation on Permanent Magnet Synchronous Machines

Machines ◽

10.3390/machines9080139 ◽

2021 ◽

Vol 9 (8) ◽

pp. 139

Author(s):

Matthias Vollat ◽

Dominik Krahe ◽

Frank Gauterin

Keyword(s):

Permanent Magnet ◽

Current Injection ◽

Synchronous Machines ◽

Permanent Magnet Synchronous Machine ◽

Permanent Magnet Synchronous Machines ◽

Harmonic Current ◽

Harmonic Currents ◽

Sensor Signals ◽

Novel Method ◽

Higher Harmonic

To reduce torque oscillations in electric motors, harmonic current injection (HCI) has been used in industry for some time. For this purpose, higher harmonic currents calculated in advance are injected into the machine. Since the general conditions for the machine can change during its life cycle, this article presents a method that makes it possible to change the parameters of HCI during operation. For this purpose, sensor signals are used to detect the reaction of the electric motor to small variations of the HCI parameters. The knowledge gained in this way is used to make further suitable variations. FEM simulations were used to verify the effectiveness of the approach. The results show that the algorithm can independently optimize the HCI parameters during runtime and reduces the amplitude of the 6th harmonic in the torque by 87% for a permanent magnet synchronous machine.

Download Full-text

Sensor Modeling for Underwater Localization Using a Particle Filter

Sensors ◽

10.3390/s21041549 ◽

2021 ◽

Vol 21 (4) ◽

pp. 1549

Author(s):

Humberto Martínez-Barberá ◽

Pablo Bernal-Polo ◽

David Herrero-Pérez

Keyword(s):

Sequential Monte Carlo ◽

Sensory Information ◽

Dead Reckoning ◽

Underwater Vehicle ◽

Position Estimation ◽

Sources Of Uncertainty ◽

Sensor Signals ◽

Underwater Navigation ◽

Underwater Localization ◽

Sensor Modeling

This paper presents a framework for processing, modeling, and fusing underwater sensor signals to provide a reliable perception for underwater localization in structured environments. Submerged sensory information is often affected by diverse sources of uncertainty that can deteriorate the positioning and tracking. By adopting uncertain modeling and multi-sensor fusion techniques, the framework can maintain a coherent representation of the environment, filtering outliers, inconsistencies in sequential observations, and useless information for positioning purposes. We evaluate the framework using cameras and range sensors for modeling uncertain features that represent the environment around the vehicle. We locate the underwater vehicle using a Sequential Monte Carlo (SMC) method initialized from the GPS location obtained on the surface. The experimental results show that the framework provides a reliable environment representation during the underwater navigation to the localization system in real-world scenarios. Besides, they evaluate the improvement of localization compared to the position estimation using reliable dead-reckoning systems.

Download Full-text

Feature Extraction from Equipment Sensor Signals with Time Series Clustering and Its Application to Defect Prediction

2020 International Symposium on Semiconductor Manufacturing (ISSM) ◽

10.1109/issm51728.2020.9377525 ◽

2020 ◽

Author(s):

Daisuke Hamaguchi ◽

Tomonari Masada ◽

Takumi Eguchi

Keyword(s):

Time Series ◽

Feature Extraction ◽

Defect Prediction ◽

Time Series Clustering ◽

Sensor Signals

Download Full-text

An Evaluation of Three Kinematic Methods for Gait Event Detection Compared to the Kinetic-Based ‘Gold Standard’

Sensors ◽

10.3390/s20185272 ◽

2020 ◽

Vol 20 (18) ◽

pp. 5272

Author(s):

Nicole Zahradka ◽

Khushboo Verma ◽

Ahad Behboodi ◽

Barry Bodt ◽

Henry Wright ◽

...

Keyword(s):

Event Detection ◽

Gold Standard ◽

Kinetic Data ◽

Force Plate ◽

Detection Methods ◽

Initial Contact ◽

Typically Developing ◽

Typically Developing Children ◽

Sensor Signals ◽

Video Sensor

Video- and sensor-based gait analysis systems are rapidly emerging for use in ‘real world’ scenarios outside of typical instrumented motion analysis laboratories. Unlike laboratory systems, such systems do not use kinetic data from force plates, rather, gait events such as initial contact (IC) and terminal contact (TC) are estimated from video and sensor signals. There are, however, detection errors inherent in kinematic gait event detection methods (GEDM) and comparative study between classic laboratory and video/sensor-based systems is warranted. For this study, three kinematic methods: coordinate based treadmill algorithm (CBTA), shank angular velocity (SK), and foot velocity algorithm (FVA) were compared to ‘gold standard’ force plate methods (GS) for determining IC and TC in adults (n = 6), typically developing children (n = 5) and children with cerebral palsy (n = 6). The root mean square error (RMSE) values for CBTA, SK, and FVA were 27.22, 47.33, and 78.41 ms, respectively. On average, GED was detected earlier in CBTA and SK (CBTA: −9.54 ± 0.66 ms, SK: −33.41 ± 0.86 ms) and delayed in FVA (21.00 ± 1.96 ms). The statistical model demonstrated insensitivity to variations in group, side, and individuals. Out of three kinematic GEDMs, SK GEDM can best be used for sensor-based gait event detection.

Download Full-text