Precision Agriculture

Northeastern North America is the world’s leading producer of wild blueberries, with over 86,000 hectares under management, and 112 million kg of fruit valued at $470 million produced annually. Wild blueberry fields are developed from native stands on deforested farmland by removing competing vegetation.

The majority of fields are situated in naturally acidic soils that are low in nutrients, have high proportions of bare spots and weed patches, and on gentle to severe topography. Traditionally, growers apply agrochemicals uniformly without considering bare spots, weed patches, or the substantial variability in soil/crop characteristics that exist within fields. Uniform application of chemicals therefore results in either over- or under-application.

Wild blueberries are low-input systems with a narrow optimal range of plant nutrients. Detrimental effects of excess N occur when too much N is applied (i.e., lowers floral bud numbers and harvestable yields). Unnecessary or over-fertilization in both bare spots and weed patches can deteriorate water quality, promote increased weed growth and reduce profit. Under-fertilization restricts yield and reduces berry quality.

The unique features of the wild blueberry cropping system as compared to other closed canopy crops emphasize the need for the development of affordable, reliable, and efficient automated precision agriculture systems for accurate applications to maximize profit and minimize environmental impacts.

Wild blueberry yields are highly dependent on agrochemicals (herbicide, fungicides, insecticide) for adequate weed, floral blights, leaf diseases, and insect control. Growers apply herbicides during the growing season (Calisto to control goldenrod, Kerb in the fall to control fescue grasses and sheep sorrel); fungicides for floral blights (Monilinia and Botrytis) and leaf diseases (Septoria and Rust); and insecticides for fruit flies. Traditionally, these agrochemicals are applied uniformly without considering significant bare spots or weed patches that exist within fields.

The repeated and excessive use of agrochemicals has resulted in increased cost of production. The overuse of agrochemicals is also dangerous for the environment, for humans, for the native pollinators, and for the plants. Chemically polluted runoff from fields causes contaminated surface and ground waters.

In order to manage wild blueberry cropping systems with reduced agrochemical inputs that have increased in diversity and are expressing increased incidence of agrochemical resistance will require: (i) increased agrochemical use efficiency; and (ii) replacement of conventional technology with new real-time VR technology for site-specific application of agrochemicals. There is an urgent need to develop affordable, reliable real-time VR applicators, using cheap sensors/cameras and controllers for spot-specific application of agrochemicals in wild blueberry cropping systems.

Wild blueberry growers enjoy high demand for their crop but also face high risks managing a product that essentially grows on uncultivated, challenging terrain. The Precision Agriculture Research Chair position evolved from meetings almost a decade ago when John Bragg, president and CEO of Oxford Frozen Foods, Ltd. met with David Percival (lead investigator and plant physiologist at the Faculty of Agriculture), prioritized research and set the goal of reducing crop protection usage in both wild blueberries and carrot production by 30%. Combined with this was the desire to maintain or improve pest control and also increase yield.

As Dr. Percival recalls, the idea was to maintain a competitive edge and sustainable, inheritable production systems while cutting costs. Producers realized that a minimalist approach was needed to retain market share and keep building product reputation. To achieve these objectives, Oxford Frozen Foods and the Faculty of Agriculture (then NSAC) started the process of establishing the Precision Agriculture Research Program (PARP) in 2002. Combined with the support of existing researchers and the wild blueberry producers associations in the Maritimes, the initial vision of the research was to substantially reduce crop protection usage, improve land stewardship and increase berry yields.  

This research was initiated in 2002 with the following objectives:

1. To test and evaluate cost-effective sensors and controllers toward developing affordable and viable precision agriculture system to maximize farm profitability and minimize environmental impacts.
2. To develop an automated cost-effective yield monitoring system using digital photography techniques to map wild blueberry fruit yield in commercial fields.
3. To develop a cost-effective automated slope sensing system to map topographic features in wild blueberry fields.
4. To evaluate soil sensing systems to quantify and map soil variation toward developing variable rate technologies in wild blueberry cropping systems.
5. To determine the feasibility of using digital geographic positioning systems/geographic information systems (DGPS/GIS) to develop prescription maps for variable rate application of agrochemicals based on the variations in soil/plant characteristics, topographic features, and fruit yield.
6. To develop a prototype variable rate spreader for site-specific fertilization.
7. To develop a prototype commercial variable rate sprayer for spot-application of agrochemicals.
8. To evaluate the impact of variable rate agrochemical applications on groundwater contamination and air quality.
9. To assess the cost/benefit of the new PA systems in blueberry production systems and eventually other high-value crops.

The outcome of the several meetings to achieve the above objectives was a six-year environmental technologies and precision agriculture initiative that received $4 million in funding.

Several researchers from the Faculty of Agriculture and other units at Dalhousie University, as well as Agriculture Agri-Food Canada worked together to achieve the objectives. Dr. Qamar Zaman joined the PARP team in December 2006 as Precision Agriculture Research Chair. 

The Precision Agriculture team placed yield-mapping research as its first priority. In 2007, digital color photography techniques were tested and evaluated to map wild blueberry fruit yields. A 10‐megapixel, 24‐bit digital color camera was mounted on a tripod and pointed downwards to take photographs of the blueberry crop from a height of approximately 1 m. At harvest time, blueberry crop images were collected for several fields at different sample locations displaying a range in yield.

Actual fruit yield was sampled from the same locations by hand‐harvesting out of a 0.5 × 0.5 m quadrant using a commercial blueberry rake. Custom image processing software was developed to count the blue pixels of ripe fruit in the quadrat region of each image and express it as a percentage of total quadrat pixels. Linear regression was used to calibrate the fruit yield with the percentage of blue pixels separately in each field. The percentage of blue pixels correlated highly significantly with hand‐harvested fruit yield (R2 = 0.98; P < 0.001). Blueberry yield maps, along with fertility and topographic maps, could be used to generate prescription maps for site-specific application of agrochemicals (Zaman et al., 2008; Transaction of ASABE Paper)

Based on these results, an automated yield monitoring system (AYMS) consisting of a digital color camera, DGPS, custom software, and a ruggedized laptop computer was developed and mounted on a specially designed farm motorized vehicle for real‐time fruit yield mapping. Wild blueberry fields were selected in central Nova Scotia to evaluate the performance of the AYMS. Ripe fruit was hand‐harvested out of a 0.5 × 0.5 m quadrant at each selected point and camera images were also taken from the same points to calculate the blue pixel ratio (percentage of blue pixels in the image). Custom software was developed to acquire and process the images in real‐time and to store the blue pixel ratio. The estimated yield per image, along with geo‐referenced coordinates, was imported into Arc GIS 9.3 for mapping (Zaman et al., 2010; Transaction of ASABE Paper)
 
After the success of the AYMS, the PA research team started working on developing a system for real-time mapping of slope (i.e., topographic features). The development of site-specific agriculture has increased the need for knowledge regarding within-field variability in factors such as soil/plant characteristics and topography that influence wild blueberry production. The majority of blueberry fields in eastern Canada have gentle to severe topography.

An automated slope measurement and mapping system (SMMS) consisting of low-cost accelerometers used as tilt sensors, DGPS and a laptop, and custom software was developed under the supervision of Dr. Qamar Zaman. The SMMS was mounted on an all-terrain vehicle for real-time slope measurement and mapping. Six commercial wild blueberry fields were surveyed in central Nova Scotia to evaluate the performance of the SMMS. The automatically sensed slopes (SS) were also compared with manually measured slopes (MS) at randomly selected points to examine the accuracy of the SMMS. The SMMS measured slope reliably in the selected fields with root mean square error ranging from 0.12 to 0.56 degrees and correlations of SS with MS of R2 = 0.95 to 0.99. The selected fields had substantial variation in slope (ranging from 0.8 to 31.0 degrees). Therefore, the use of low-cost and reliable accelerometers with a DGPS is a better option than expensive real-time kinematic DGPS for developing cost-effective SMMS to quantify and map slopes (real-time) for planning site-specific management practices in commercial fields. The SS maps or real-time SMMS could also be used to adjust vehicle speed at particularly steep slopes (Zaman et al., 2010; Hort. Technology paper).

After the success of the SMMS, Dr. Zaman and his team started working on evaluating the electromagnetic induction (EMI) sensor for mapping soil properties in wild blueberry fields. EMI methods are gaining popularity due to their non-destructive nature, rapid response, and ease of integration into mobile platforms for assessment of soil moisture content, water table depth, salinity, etc. Detailed geo-referenced maps would be useful for site-specific management of agricultural inputs. The ground conductivity readings were recorded with Dual EM (DualEM, Milton, Ontario, Canada) at sleeted grid points.

Soil samples were also collected from the same sampling points and analyzed for selected soil parameters using standard methods. The ground conductivity readings were significantly correlated with the selected soil properties (soil organic matter [SOM], moisture content, soil texture). Wells were also installed in wild blueberry fields to estimate the water table depth using EMI. Comprehensive surveys were conducted in the selected fields to measure ground conductivity for soil properties and water table depth estimation in real-time using DualEM and DGPS (Farooque, 2011; MSc thesis).

The majority of wild blueberry fields have high proportions of bare spots and weed patches (from 30 to 50%). Producers presently apply agrochemicals uniformly without considering bare spots, which may result in unnecessary or over‐application of agrochemicals in bare spots that in turn may increase environmental pollution and the cost of production. An automated, cost‐effective machine vision system using digital color photography was developed and tested to detect and map bare spots for site‐specific application of agrochemicals within wild blueberry fields.

The machine vision system, consisting of a digital color camera, DGPS, and a notebook computer, was mounted on a specialized farm vehicle. Custom software for capturing and processing color images was developed in Delphi 5.0 and C++ programming languages. The images taken by the digital camera were stored in the notebook computer automatically and then processed in red, green, and blue (RGB), and hue, saturation, and value (HSV) color spaces to detect bare spots in real‐time within blueberry fields. The best results were achieved in hue image color space with 99% accuracy and a processing speed of 661 ms per image. The results indicated that bare spots could be identified and mapped with this cost‐effective digital photography technique in wild blueberry fields. This information is useful for site‐specific application, and has the potential to reduce agrochemical usage and associated environmental impacts in the wild blueberry production system (Zhang et al., 2010 and Chang et al., 2011, Applied Engg. Agri. paper).

The unique features of these technologies function significantly better than existing precision agriculture systems, and at lower cost. This type of VR sprayer will not use prescription maps, but will rely on sensors/digital cameras to provide real-time weed, bare spots and blueberry plant detection information that will be used to dispense correct agrochemical rates site-specifically within wild blueberry fields. Most importantly, the herbicide will not be applied where no weeds have been detected and fungicides/insecticides will be applied only where blueberry plants have been detected, thus reducing waste and contamination to the environment.

In 2009, an automated prototype VR sprayer was developed for control of 8 individual nozzles on a 6.1 m sprayer boom for in-season, site-specific application of agrochemicals on weeds. The sprayer boom was divided into 8 sections and mounted behind an all-terrain vehicle (ATV) at 76.2 cm above the ground. The variable-rate control system consisted of 8 ultrasonic sensors (one per spray section) mounted on a separate boom in front of the ATV, DICKEY-john Land Manager II controller and flow valve, solenoid valves and an 8-channel VR controller interfaced to a Pocket PC (PPC) using wireless Bluetooth® radio with Windows Mobile® compatible software. This type of VR sprayer does not use prescription maps, but relies on sensors to provide real-time weed detection information that is used to dispense correct agrochemical rates for the weeds in a specific boom section where the weeds have been detected.

Not all weeds are tall in the wild blueberry cropping system; to identify small weeds, color contrast and/or textural feature-based software was developed and connected with the VR controller. In 2010, a prototype was created: a 20-foot sprayer equipped with 2 µEye digital color cameras on each side of the front boom, which are attached by USB cables to the computer. Custom software processes the images to detect weeds, bare spots and blueberry plants in real-time. This device has user-programmable inputs, such as a before-and-after buffer, time delay, and ground speed corrections. In turn, these weed- or plant-triggering signals are transmitted to the 8-channel computerized VR controller. In effect, the flow rate can be adjusted automatically based on the number of nozzles operating at a specific time. The corresponding solenoid sprayer valve could be activated to spray crop protection product in the specific boom section where the weeds or plants have been detected. The vision of the research was to substantially reduce agrochemical usage, improve land stewardship, protect the environment, and increase berry yields. (Zaman et al., 2011: Computer and Electronics in Agriculture paper).

In fall 2010, a prototype commercial (45-foot boom) VR sprayer was developed. The 45 ft. MS 1135E sprayer boom is divided into sixteen sections and mounted behind a John Deere 6430 farm tractor. Sixteen solenoid valves and nozzles are installed on the rear boom and connected with the 8 channels computerized controllers. Eight cameras (one for every two sections) are incorporated vertically on a 45 ft front mounted boom and attached by USB cables to the computer in the tractor cabin. Two 8-channel computer controllers receive triggering signals from the custom-made image processing software. The corresponding solenoid sprayer valve is activated to spray crop protection product in the specific boom section where the weeds or plants have been detected. Computerized controllers communicate automatically to the 20-channel MidTech controller, which regulates the flow rate to the nozzles through a servo valve. A self-leveling sensing system was introduced in front boom to adjust camera heights automatically during field operation. In addition to the substantial crop protection savings, the researchers considered the environmental benefits of a crop that must maintain its wild and pristine image for markets in Japan and Europe. Preliminary results indicate cost savings of up to 80%.

Present nutrient management recommendations are typically uniform without considering the variation in soil and plant characteristics. However, the soils are highly variable spatially and, therefore, the uniform management of agricultural inputs may results in over-application in areas with high productivity and under-application in areas with low productivity. Site-specific management of nutrients using VR spreader has been acknowledged as one means of addressing this problem. The most popular approach to manage spatial variability within fields is the use of management zones (MZs), in which field that have relatively homogeneous attributes in landscape and soil condition are subdivided, and this technique can be used to direct variable rate fertilizer application. The wild blueberry field has been reported to have significant bare spots and weed patches. Over-fertilization in weed patches and bare spots with conventional methods may affect water quality, and increase production cost. Therefore there is a need to develop map/sensor based variable rate spreader for accurate site-specific fertilization to maximize profit and minimize environmental impacts. The fertilizer spreader consists of a controller, metering device, DGPS, laptop computer, and Site Mate Farm Works Software. The variable rate spreader has been tested and evaluated for performance accuracy in wild blueberry fields. Repentantly, a prescription map was generated in ArcGIS based on the slope variation. The bare spots were mapped using GPS and were defined as a separate class in the prescription map. Zero fertilizer was allocated in the bare spots. One part of the field was fertilized uniformly at grower’s rate. The prescription map was uploaded in Farm Works software to apply variable rate fertilization as prescribed in the map. Lysemeters were also installed in the experimental field. The purpose of this experiment was to visualize the impact of variable fertilization on fruit yield and ground water quality.

Currently, a project is underway to evaluate the performance accuracy of the commercial variable rate sprayer and spreader, their economic analysis and environment benefits. The real-time kinematic GPS has been under use for mapping weeds and bare spots to examine the detection of weeds and chemical overlap.

• Automated Yield Monitoring System
• Automated Slope Mapping System
• Soil Sensing Using Electromagnetic Induction Method
• Automated Prototype Variable Rate Sprayer for Spot-Application of Agrochemicals
• Automated Prototype Commercial Variable Rate Sprayer
• Variable Rate Granular Fertilizer Spreader
• Spot-Application of Fertilizer Using Automated Sensing and Control System (Phase-II)
• Machine Vision System for Spot-Application of Agrochemicals in Wild Blueberry Fields
• Environmental Impact of VRT on Ground Water Contamination
• Economic Analysis of Precision Agriculture Systems
• Cost Analysis-Conventional Vs. Spot-Application (One application only)

Equipment

We have numerous images of our equipment being used in research projects out in the field. Please contact us if you would like to see some, or if you would like to know more about our research.

Maps

We also have various toporaphical and other maps available. Please contact us if you would like to see these maps, or if you have any questions about our research.

Precision Agriculture (PA) covers a research area with goals to optimize agricultural production systems in both time and spatial dimensions. The concept of PA technologies has been proposed as a solution to manage spatial and temporal variability to more efficiently apply agricultural inputs for the purpose of improving crop performance and environmental quality.

In practice, PA changes the way a farmer works:

• Fruit yields are not only harvested but also mapped using a combination of sensors, digital photography techniques, and DGPS/GIS.
• Soil sensing systems provide information on the variability in soil productivity status.
• Crop sensing technology provides information about canopy characteristics.
• Fertilizers are allocated more efficiently by exploiting spatial variations in soil fertility levels according to local demand.
• Agrochemicals can be sprayed site-specifically on an as-needed basis.
• New techniques, such as variable rate technologies (VRT) including sprayers, spreaders, sensors, controllers, electromagnetic induction (EMI) method, DGPS, GIS, ultrasonic sensor systems, remote sensing, etc., can be used.
• Information technology supports decision making in the field based on the information collected via PA technologies.

Dr. Qamar U. Zaman
Associate Professor/Precision Agriculture Research Chair
Department of Engineering
Precision Agriculture Research Program
Faculty of Agriculture, Dalhousie University
Truro, NS, Canada  B2N 5E3
Phone: (902) 893-5426
Fax: (902) 893-1859
Email: qzaman@nsac.ca