LoRa Based IoT Platform for Remote Monitoring of Large-Scale Agriculture Farms in Chile.

Nowadays, conventional agriculture farms lack high-level automated management due to the limited number of installed sensor nodes and measuring devices. Recent progress of the Internet of Things (IoT) technologies will play an essential role in future smart farming by enabling automated operations with minimum human intervention. The main objective of this work is to design and implement a flexible IoT-based platform for remote monitoring of agriculture farms of different scales, enabling continuous data collection from various IoT devices (sensors, actuators, meteorological masts, and drones). Such data will be available for end-users to improve decision-making and for training and validating advanced prediction algorithms. Unlike related works that concentrate on specific applications or evaluate technical aspects of specific layers of the IoT stack, this work considers a versatile approach and technical aspects at four layers: farm perception layer, sensors and actuators layer, communication layer, and application layer. The proposed solutions have been designed, implemented, and assessed for remote monitoring of plants, soil, and environmental conditions based on LoRaWAN technology. Results collected through both simulation and experimental validation show that the platform can be used to obtain valuable analytics of real-time monitoring that enable decisions and actions such as, for example, controlling the irrigation system or generating alarms. The contribution of this article relies on proposing a flexible hardware and software platform oriented on monitoring agriculture farms of different scales, based on LoRaWAN technology. Even though previous work can be found using similar technologies, they focus on specific applications or evaluate technical aspects of specific layers of the IoT stack.

Three-wavelength broadband soot pyrometry technique for axisymmetric flames.

Soot temperature measurements in laminar flames are often performed through two-color broadband emission pyrometry (BEMI) or modulated absorption/emission (BMAE) techniques, using models to relate the ratio between flame intensities at two different wavelengths with soot temperature. To benefit from wider spectral range and increase the accuracy of experimental estimation of soot temperature, this work proposes a new approach that uses three-color broadband images captured with a basic color camera. The methodology is first validated through simulations using numerically generated flames from the CoFlame code and then used to retrieve soot temperature in an experimental campaign. The experimental results show that using three-color and BEMI provides smoother reconstruction of soot temperature than two-color and BMAE when small disturbances exist in the measured signals due to a reduced experimental noise effect. A sensitivity analysis shows that the retrieved temperature from three-color BEMI is more resilient to variations on the ratio of measured signals than BMAE, which is confirmed by an error propagation analysis based on a Monte Carlo approach.

Analysis of Soot Propensity in Combustion Processes Using Optical Sensors and Video Magnification.

Industrial combustion processes are an important source of particulate matter, causing significant pollution problems that affect human health, and are a major contributor to global warming. The most common method for analyzing the soot emission propensity in flames is the Smoke Point Height (SPH) analysis, which relates the fuel flow rate to a critical flame height at which soot particles begin to leave the reactive zone through the tip of the flame. The SPH and is marked by morphological changes on the flame tip. SPH analysis is normally done through flame observations with the naked eye, leading to high bias. Other techniques are more accurate, but are not practical to implement in industrial settings, such as the Line Of Sight Attenuation (LOSA), which obtains soot volume fractions within the flame from the attenuation of a laser beam. We propose the use of Video Magnification techniques to detect the flame morphological changes and thus determine the SPH minimizing observation bias. We have applied for the first time Eulerian Video Magnification (EVM) and Phase-based Video Magnification (PVM) on an ethylene laminar diffusion flame. The results were compared with LOSA measurements, and indicate that EVM is the most accurate method for SPH determination.

The effect of a lytic polysaccharide monooxygenase and a xylanase from Gloeophyllum trabeum on the enzymatic hydrolysis of lignocellulosic residues using a commercial cellulase.

Hydrolysis of lignocellulosic biomass depends on the concerted actions of cellulases and accessory proteins. In this work we examined the combined action of two auxiliary proteins from the brown rot fungus Gloeophyllum trabeum: a family AA9 lytic polysaccharide monooxygenase (GtLPMO) and a GH10 xylanase (GtXyn10A). The enzymes were produced in the heterologous host Pichia pastoris. In the presence of an electron source, GtLPMO increased the activity of a commercial cellulase on filter paper, and the xylanase activity of GtXyn10A on beechwood xylan. Mixtures of GtLPMO, GtXyn10A and Celluclast 1.5L were used for hydrolysis of pretreated wheat straw. Results showed that a mixture of 60% Celluclast 1.5L, 20% GtXyn10A and 20% GtLPMO increased total reducing sugar production by 54%, while the conversions of glucan to glucose and xylan to xylose were increased by 40 and 57%, respectively. This suggests that GtLPMO can contribute to lignocellulose hydrolysis, not only by oxidative activity on glycosidic bonds, but also to hemicellulose through the oxidation of xylosyl bonds in xylan. The concerted action of these auxiliary enzymes may significantly improve large-scale recovery of sugars from lignocellulose.

Model, analysis, and evaluation of the effects of analog VLSI arithmetic on linear subspace-based image recognition.

Typical image recognition systems operate in two stages: feature extraction to reduce the dimensionality of the input space, and classification based on the extracted features. Analog Very Large Scale Integration (VLSI) is an attractive technology to achieve compact and low-power implementations of these computationally intensive tasks for portable embedded devices. However, device mismatch limits the resolution of the circuits fabricated with this technology. Traditional layout techniques to reduce the mismatch aim to increase the resolution at the transistor level, without considering the intended application. Relating mismatch parameters to specific effects in the application level would allow designers to apply focalized mismatch compensation techniques according to predefined performance/cost tradeoffs. This paper models, analyzes, and evaluates the effects of mismatched analog arithmetic in both feature extraction and classification circuits. For the feature extraction, we propose analog adaptive linear combiners with on-chip learning for both Least Mean Square (LMS) and Generalized Hebbian Algorithm (GHA). Using mathematical abstractions of analog circuits, we identify mismatch parameters that are naturally compensated during the learning process, and propose cost-effective guidelines to reduce the effect of the rest. For the classification, we derive analog models for the circuits necessary to implement Nearest Neighbor (NN) approach and Radial Basis Function (RBF) networks, and use them to emulate analog classifiers with standard databases of face and hand-writing digits. Formal analysis and experiments show how we can exploit adaptive structures and properties of the input space to compensate the effects of device mismatch at the application level, thus reducing the design overhead of traditional layout techniques. Results are also directly extensible to multiple application domains using linear subspace methods.

Analysis and compensation of the effects of analog VLSI arithmetic on the LMS algorithm.

Analog very large scale integration implementations of neural networks can compute using a fraction of the size and power required by their digital counterparts. However, intrinsic limitations of analog hardware, such as device mismatch, charge leakage, and noise, reduce the accuracy of analog arithmetic circuits, degrading the performance of large-scale adaptive systems. In this paper, we present a detailed mathematical analysis that relates different parameters of the hardware limitations to specific effects on the convergence properties of linear perceptrons trained with the least-mean-square (LMS) algorithm. Using this analysis, we derive design guidelines and introduce simple on-chip calibration techniques to improve the accuracy of analog neural networks with a small cost in die area and power dissipation. We validate our analysis by evaluating the performance of a mixed-signal complementary metal-oxide-semiconductor implementation of a 32-input perceptron trained with LMS.