SBIR-STTR Award

The use of GPS-aided technology to apply inputs site-specifically has a significant potential to improve the use of fertilizers, lime, and other materials. One of the main factors limiting the effectiveness of this technology is the problem of accurately assessing the input requirements of the field. In the most common approach, soil samples are acquired on a 2.5 acre grid pattern, sent to a laboratory for analysis, and various interpolation methods are applied to the sample results in order to predict the values for the entire field. Studies have shown that soil properties often have spatial variability greater than 2.5 acres, however the costs of sampling and lab analysis preclude denser sampling. A promising approach to this problem is to use devices that measure soil properties on-the-go more intensively than 2.5 acres/samples. One such sensor was prototyped by Purdue University for sensing soil pH. It demonstrated that soil pH could be successfully measured in-situ using combination, ion-selective electrodes. For this device to be viable, research must be conducted to create a soil sampling mechanism that can acquire soil in a wide variety of soil conditions and at commercially acceptable operating speeds.

Anticipated Results/Potential Commercial Applications of Research:
Currently, in excess of $10 billion of fertilizers and lime are applied to U.S. fields (NASS). Growers are becoming increasingly aware of the variability within these fields, yet they also recognize that spatial variabilities of properties may exceed what 2.5 acre grid sampling can identify. Many are waiting to adopt precision practices until improvements are made in the technology. Once a soil-sampling mechanism is developed, on-the-go pH measuring can become a reality. Research into the feasibility of using combination, ion-selective electrodes to measure potassium and nitrogen levels in situ is currently underway, which would require a similar sampling mechanism.

As crop growers begin to use GPS technology to manage the variability in their fields, they are seeking ways to assess the variability of physical and chemical properties more accurately and cost-effectively. In the most common approach, soil samples are acquired on a 2.5 acre grid pattern. Soil properties often have greater spatial variability than is identified with this approach, however the costs of sampling and lab analysis preclude denser sampling. A promising approach to this problem is to use ion-selective electrodes that measure soil properties on-the-go much more intensively than 2.5 acres/sample. A soil-sampling mechanism was prototyped by Purdue University for sensing soil pH. This device was re-designed and extensively field-tested with pH electrodes during a Phase I project. On-the-go pH sensing has proven feasible and is viable for commercial development. Validation of pH sensor data during earlier research confirmed the economic advantages of dense, on-the-go mapping versus current sampling methods. In the new project, several user-oriented features will be integrated for commercial use, and advancements will be made in using other ion-selective electrodes, and in prescribing applications from sensor data. OBJECTIVES: The first objective for this project will be to construct a new prototype soil pH sampling system with multiple sensor modules and EC electrodes. Here are some specific objectives for this system: 1) hydraulic power will come from a small gasoline engine powering a hydraulic pump, 2) wheel and electrode spacing will allow mapping growing 30" row crops, 3) operating cost of electrodes will be improved through breakage resistant design, 4) electrode holder will be redesigned, making it easier to calibrate, remove, store, and change electrodes, 5) controller module controlling of automated cycling functions of multiple electrodes will be developed, 6) Veris instrument and/or field computer will measure soil EC simultaneously with logging ISE data, 7) operating software will be written that reviews sensor data during data collection and interacts with controller module, 8) software will be written that will recognize when the electrodes have stabilized, according to predetermined criteria, 9) a warning function will be added that notifies the operator whenever sensor response indicates a possible electrode or mechanism failure, 10) pH calibration procedure will be added, and 11) validation of the new automated system. The second set of objectives relates to additional electrode capability. The sampling system will be outfitted with K, N, Na electrodes, and a comprehensive set of field tests will be conducted. The sensor data will be validated with lab analyzed soil samples, the durability and reliability of electrodes in lab and in the field will be evaluated, and operational issues such as calibration, electrode cleaning, and controls will be optimized. The third set of objectives is for deriving prescriptive information from on-the-go mapping using optical and EC data along with ion-selective-electrode mapping. This entails using ion-selective electrodes to map the spatial variability in pH, K, and N, and deriving a lime, potash, and nitrogen application information using optical and EC data along with advanced geo-statistical techniques. The accuracy performance criteria for all testing will involve comparing the results of this automated system to the results of conventional and grid sampling. The durability criteria are similar-comparing the cost of using the automated system with grid sampling. APPROACH: Two identical new prototypes will be designed and constructed. They will include two alternately cycling ion-selective-electrode modules, and six EC coulter-electrodes spaced to avoid damage to row crops. These will be pull-type implements for use with either tractor or 4WD pickup truck. These will feature continued improvements in electrode durability, cleaning and holding. The existing controller module will be redesigned to control automated cycling functions of alternately cycling ion-selective-electrode modules. Software will be created for the Veris instrument that will monitor data from the ion-selective-electrode, and warn operator of any data collection problems, based on criteria such as electrode stabilization, or change in reading from previous sample. A pH calibration procedure will be added to the system, where the electrodes are brought in contact with a buffer 4 and buffer 7 (and/or 10) solution, the operator initiates a calibration mode, and the software adjusts data collected since last calibration. Additional calibration methods will also be considered, including monitoring the pH during the washing cycle and calibrating based on electrode performance during washing. For validation of the ion-selective data, several fields with contrasting soils and conditions will be mapped, sampled, lab-analyzed, and the results compared to the sensed data. Validation of automated sensing will be done in regions of the US where the property of interest-pH, potassium, nitrogen, and sodium are considered important by local consultants and growers. Sampling fields for validation will be done using both grids and targeted samples. First, the field will be grid sampled on a basis that most closely approximates typical commercial sampling: 1 acre and/or 2.5 acre grids, with 8-10 cores in a 15-20' radius around the center point of grid, with the center point geo-referenced with DGPS. On each field, 5-10 targeted calibration samples per field will be chosen within areas that showed spatially structured pH. It has been previously shown that water used in the field affects electrode output. Distilled and deionized water are used in a laboratory setting to reduce this effect. However, potential customers for the automated soil mapping system will wish to use alternative water sources. Therefore, a study will be conducted to quantify the effects of various water sources and to provide a recommendation regarding the usage of non-purified water. Laboratory and field data obtained through this task will be used to develop recommendations on appropriate utilization of the commercialized automated pH system. Geostatisical, agronomic, and economic evaluations will be combined to demonstrate the effects of various mapping densities on errors, and to compare the value of information obtained with the mapping cost