In this paper, we presented the development of a proximal soil moisture sensor that measured the soil moisture content of dairy pasture directly from the boom of an irrigator. The proposed sensor was capable of soil moisture measurements at an accuracy of +/- 5% volumetric moisture content, and at meter scale ground area resolutions. The sensor adopted techniques from the ultra-wideband radar to enable measurements of ground reflection at resolutions that are smaller than the antenna beamwidth of the sensor. An experimental prototype was developed for field measurements. Extensive field measurements using the developed prototype were conducted on grass pasture at different ground conditions to validate the accuracy of the sensor in performing soil moisture measurements.

Ground-based robots for agricultural applications are receiving increased attention as labour availability and cost drive interest and uptake. Here we describe a radar method that is suitable for soil moisture measurement from a moving vehicle. We have employed techniques from impulse radar that enable measurement of soil reflection at resolutions smaller than the antenna beam width. We show how the resolution is affected by the radar pulse-width, antenna parameters, radar location and sensing angle. The radar backscatter coefficient is calculated from the radar signals themselves, and demonstrated by measurements taken on wet pasture at different elevation angles. Measurements show that radar backscatter coefficients can be reliably measured at angles up to 60 degrees from the nadir.

Abstract This paper describes an empirical transform between the propagation time (tp) data obtained from a non invasive Time Domain Reflectometry (TDR) sensor to the percentage moisture content θv within two different road making basecourse aggregate material. Results show that a simple quadratic fit between tp and θv can be given leading to a maximum error in the estimate of 0.55%. It is also shown that the dielectric model underlying each of the basecourses is different enough to warrant the use of a unique quadratic function (i.e. different quadratic coefficients) for each.

A proof-of-concept proximal soil moisture sensor is developed for the precision irrigation of dairy pasture. The proposed sensor, mounted on the boom of an irrigator, measures the soil moisture of ground area in front of the irrigator and feeds the measured moisture data to the variable rate irrigator's controller. The sensor adopts techniques from the ultrawideband radar, in which the ground's resolution area is confined by both its antenna beamwidth and transmitted pulse-width. In this paper, we present the sensing methodology, design, prototyping, measurement and results of the proposed sensor. Measured results demonstrate the sensor's performance in terms of its soil moisture measurement accuracy, and the effects of varying sward heights of pasture grass.

We present the design of an antenna array for a proximal soil moisture sensor that will be mounted on an irrigator. The sensor maps the soil moisture of the ground ahead and uses the information to modulate the water volume of the irrigator. The design objective of the antenna is to enable the sensor to map the ground surface at less than 1m2 resolutions. Applying radar analysis, it is determined that the antenna needs to maintain a directivity performance of 12 dBi over a bandwidth of 500 MHz. An array of four log-periodic dipole antennas is designed, constructed and measured. Measurement showed that the array exhibit a flat gain performance of 10.3 dBi over the frequency range of 400-1200 MHz. The dimensions of the antenna array are 0.35m×0.45m×0.55m. Based on the measured parameters, we computed that the sensor is capable of achieving a ground resolution of 0.62m2.

Dielectric models are used with permittivity measurements of material for translation from permittivity to moisture content. A dielectric model for pure sand was developed based on fundamental physical properties such as the permittivity and geometry of the host material particles, and the frequency dependent processes that determine the permittivity of water. The measurement of sand using the short-circuited reflection method is discussed included associated pre-processing as is the processing of the measured data to extract permittivity values. A system is derived to automate the selection of appropriate solution equation, time-gate position and generation of initial values for numerical inversion. Measured data in the 1 - 6 GHz frequency range for sand with various volumetric moisture contents is compared with the dielectric model.

A major challenge with RFID tags is to obtain their coordinates in the cost effective manner that makes them so attractive in the first place. For example, adapting well known radar methods for coordinate registration will increase the RFID system complexity and thus cost and maintenance considerably. In this paper we develop a system to determine the location of RFID tags using RSS (Received Signal Strength) measurements between tags and the reader to estimate their position. Tag positioning with this system can be made with a single portable reader without the need for triangulation. The WSN (Wireless Sensor Network) is treated as an optimisation problem where relative positioning is found using a MCMC (Markov Chain Monte Carlo) technique. Simulations show that using this process it is possible to improve estimates for tag location at long ranges without major modification to currently available systems.

The determination of moisture content in raw timber is an important parameter to determine its processing life cycle. In this paper two novel methods are presented to measure this, one using the radiant electric field from a pair of parallel transmission lines and the other the capacitance built up between contact electrodes. Both methods are arranged to provide a tomographic solution to the distribution to improve either resolution (parallel line) or volume coverage (capacitance). It is shown that while improvements can be made to each technique both have the potential to accurately provide a measure of the required parameters.

We are developing a method that allows remote soil moisture measurement from electromagnetic sensors that are mounted on irrigators. To do that, techniques from the impulse (ultra-wideband) radar are applied, which enables the measurement of soil reflection at resolutions smaller than the antenna beamwidths of the radar. We derive a relationship between the radar's parameter (i.e. pulse-width, antenna parameters, radar location and sensing angle with respect to ground level) and the soil surface resolution that the radar is capable of achieving. In addition, we also propose a method to calculate the radar backscatter coefficient of the ground from the signals that are received by the radar. Using the proposed method, radar measurements are performed on wet pasture to measure the radar backscatter coefficients at different grazing angles. Measured results show that radar backscatter coefficients can be reliably measured at the angles up to 60 degrees from nadir. This paper presents the progress in our ongoing work to develop a small and practical remote soil moisture measurement for smart irrigation.

Dielectric models are used with permittivity measurements of material for translation from permittivity to moisture content. A dielectric model for pure sand was developed based on fundamental physical properties such as the permittivity and geometry of the host material particles, and the frequency dependent processes that determine the permittivity of water. The measurement of sand using the short-circuited reflection method is discussed as is the processing of the measured data to extract permittivity values. Measured data in the 1-6 GHz frequency range for sand with a volumetric moisture content of 6.35 and 11.47% and thickness of 100 and 200mm is compared with the dielectric model.