AI-Driven Autonomous Vehicle Systems for RealTime Crop Mapping and Variable Rate Fertilization

Abstract. Wild blueberry fields have gentle to severe topography with high proportion of bare spots, and weed patches emphasizing the need for variable rate (VR) fertilization. To evaluate the impact of VR fertilization on subsurface water contamination, two wild blueberry fields were selected in Nova Scotia, Canada. The field boundaries, weeds and bare spots were mapped using real-time kinematics global positioning system. The variation in slope was mapped and management zones i.e. steep slope (Zone-1), moderate slope (Zone-2) and low lying areas (Zone-3) were delineated. Both fields were divided into two halves (VR and uniform). The GPS-guided prescription maps were generated in ArcGIS 9.3 for VR treatment sections. Three fertilization rates 200, 150, and 100 kg ha -1 were applied in Zone-1, Zone-2 and Zone-3, respectively, with a VR fertilizer spreader. The bare spots/weeds were classified as a separate class in the prescription maps and no fertilizer was applied in those areas. For comparison, the uniform fertilizer rate of 200 kg ha -1 used by growers was applied in uniform treatment sections of the selected fields. Subsurface water samples were collected after every heavy rainfall (>20 mm) from 36 suction lysimeters placed in each field with 18 lysimeters in each treatment. The subsurface water samples were analyzed for nitrate nitrogen (NO 3 — N), and ammonium nitrogen (NH 4 + -N) concentrations under VR and uniform fertilization. Leaf samples were collected from both VR and uniform treatments and were analyzed for leaf nutrient concentrations. The results of this study suggested significantly lower NO 3 — N and NH 4 + -N loading from subsurface water samples leaving the root zone in moderate and low lying areas for VR treatment as compared to uniform treatment. The nitrogen, phosphorus, and potassium nutrient levels in leaves were found to be higher than the recommended ranges for uniform treatment as compare to VR treatment suggesting the excess uptake by plants in uniform treatment which may result in more vegetative growth and lower yield. The results indicated that VR fertilization based slope can increase input use efficiency, farm profitability and reduce subsurface water contamination.

HighlightsThe influencing factors of the fertilizer dropping time are determined by simulation.The relationship between the influencing factors and the fertilizer dropping time is explored on the VRF applicator.A new method is established to calculate the fertilizer dropping time.Abstract. Lag time is one of the main factors that affect the accuracy of the variable rate fertilizer (VRF) applicator. The key to correcting the impact of the lag time is to determine its influencing factors. This article focuses on the fertilizer dropping time, which is a part of the lag time, and hypothesizes that the rotational speed of the fluted roller, working length of the fluted roller, number of fertilizer particles, and the length of the fertilizer tube may affect the fertilizer dropping time. First, taking advantage of the simple and convenient setting of the working parameters and conditions of the simulation software, different working parameters were set by the EDEM software for simulation, and the simulation results were analyzed to determine that the rotational speed of the fluted roller and the length of the fertilizer tube are the main influencing factors of the fertilizer dropping time. Second, to explore the relationship between the two factors and the fertilizer dropping time in the actual work, experiments were performed on the VRF applicator. The experimental results are analyzed by ANOVA, and the analysis results show that: the correlation coefficient R2 between the rotational speed of the fluted roller and fertilizer dropping time is greater than 0.95, and there is a linear relationship between the factor and fertilizer dropping time; the fertilizer dropping time increases with increasing the length of the fertilizer tube. Finally, the rotation speed-the length of the fertilizer tube-fertilizer dropping time model was established through the multiple regression analysis. With the model, the fertilizer dropping time was accurately calculated according to the rotational speed of the fluted roller and the length of the fertilizer tube. The accuracy of the model is greater than 85%, which helps to calculate the fertilizer dropping time. This study helps to correct the lag distance by sending instructions in advance and improving the accuracy of the VRF applicator. Keywords: Fertilizer dropping time, Fertilizer tube, Rotation speed, VRF.

According to the operator experiences, the variable rate fertilization system based on the fuzzy self-tuning PID controller is researched in this paper, which is driven by electrically and hydraulic system. This system solves the problem that the parameters adjustment of the traditional PID is difficult. The system model is presented and is simulated by using MATLAB/ SIMULINK toolbox. The simulation results shows that the fuzzy self-tuning PID control could obtain better dynamic response and achieve the expected control effects. In order to verify this control system performance, many experiments are carried on in the lab. The experiments results show that the variable rate fertilization system based on fuzzy self-tuning PID control has high control accuracy and could meet the control requirements of the variable rate fertilization in precision agriculture.

Variable rate fertilizer application has the potential to improve fertilizer use efficiency, increase economic returns,and reduce environmental impacts. This study was designed to examine differences in yield and economic returns betweenuniform and variable rate fertilizer applications. During the 1997 and 1998 growing seasons, a variable rate applicator,capable of varying two liquid fertilizers simultaneously, was used to evaluate three fertility strategies: conventional uniformN, uniform N and P, and variable rate N and P. The three treatments were assigned in six blocks within three 14ha grainsorghum fields (two blocks in each field) in a randomized complete block design. Thirtysix soil samples were taken in astaggered systematic grid from each field, and levels of soil nutrients were determined. Application rate maps for the variablerate N and P treatment were generated based on a fixed yield goal and sitespecific soil N and P levels across the experimentalplots, while application rates for the uniform N and P treatment were calculated from the same yield goal and average soilN and P levels for all three fields. Yield monitor data indicated that the variable rate treatment resulted in significantly higheryields than the uniform N and P treatment for both years (400 kg/ha higher in 1997 and 338 kg/ha higher in 1998). Moreover,coefficients of variation of yield monitor data for the variable rate treatment were smaller than those for the two uniform ratetreatments. A simple economic analysis showed that the variable rate treatment had positive relative economic returns overthe uniform N and P treatment ($27/ha in 1997 and $23/ha in 1998). However, if additional costs for soil sampling, equipment,and data analysis associated with variable rate application were considered, these returns would be much lower or evennegative. These results showed that variable rate fertilization can increase yield, reduce yield variability, and improveeconomic returns. More experiments are needed to evaluate the longterm agronomic, economic, and environmental viabilityof variable rate technology in the Rio Grande Valley of south Texas

Site-specific farming (SSF) practices are being adopted at an accelerating rate, but evidence of their profitability has been mixed or missing. This contribution evaluates the profitability of SSF practices by synthesizing quantitative and qualitative research results within the context of the economics of information technology. The profitability results from nine published field research studies on variable rate (VR) fertilizer application are reviewed using partial budgets adjusted to include minimum costs and grid cell areas for each study. Profitability results correlated closely with the gross revenue earning potential of the crop, so VR fertilizer application was unprofitable on wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.), sometimes profitable on corn (Zea mays L.), and profitable on sugarbeet (Beta vulgaris subsp. L. vulgaris). Although the formal published literature has ignored the profitability of yield mapping, production economics and farmer interviews suggest that yield mapping is profitable when it reveals yield patterns that can be managed at acceptable cost and when the information has compensating off-field value. Manageable yield variability includes not only VR input management, but also whole-field improvements such as field drainage, land leveling, windbreaks, and fencing. Off-field value can come from cheaper on-farm experimentation and greater negotiating power with landlords. Farmers and agribusinesses committed to field crop production for the long term should develop SSF capabilities. But because SSF practices are site-specific, their profitability potential too is site-specific. This site specificity extends beyond the farm field to the crop rotation, and the capabilities and opportunities available to the farm or agribusiness manager.

High-precision and low-cost electronic diagnostic devices are needed for monitoring crop nitrogen status and variable rate fertilization. One normalized difference vegetation index portable sensing device was designed in this study based on the peak red wavelength of 650 nm and the peak near-infrared wavelength of 940 nm. A revised normalized difference vegetation index formula was developed based on optical absorption and transmission principles and through extensive experiments, and a prototype of the device with both hardware and software was constructed. Comparison experiments using corn leaves showed that the device provided comparable measurement results with a SPAD-502 meter. The revised normalized difference vegetation index formula proposed in this study will be useful for other applications and the portable sensing device designed will provide a useful tool for real-time, non-damage monitoring of crop nitrogen status and for the implementation of variable rate fertilization in precision agriculture.

We describe the use of machine learning methods in the analysis of spatial soilfertility, soil physical characteristics, and yield data, with a particular objective of determininglocal (field- to farm-scale) crop response patterns. For effective prescriptive use, the output ofthese tools is augmented with economic data and operational constraints, and recast as a rule-baseddecision support tool to maximize economic return in variable rate fertilization systems.We describe some of the practical issues addressed in development of one such system,including data preparation, adaptation of regression tree output for use in a rule-based expertsystem, and incorporation of real-world limits on system recommendations. Results fromvarious field trials of this system are summarized.

Orchard target-oriented fertilizing with on-board sensor technology can improve fertilizer efficiency and reduce environmental pollution. Photoelectric sensors are widely used for object detection because of their low cost and fast response time. This paper presents a Wolfberry tree dual-model detection method and the design of an orchard target-oriented variable-rate fertilization system based on photoelectric sensors. The dual-model detection method includes the Trunk Detection Model (TDM) and Canopy Detection Model (CDM), which can be applied for Wolfberry orchards at the green cluster and mature stages, respectively. A target-oriented fertilization system was designed using the dual-model method, and tested in the lab and Chinese Wolfberry orchard. The laboratory test results showed that the average center offset distances on the condition of detecting trunks, continuous canopies, and discontinuous canopies were 4.1 cm, 9.1 cm and 13.1 cm, respectively. The system could ignore the signals from canes when their diameters were less than 16 mm, and also could determine the gaps within a tree when they were less than 21 cm. The orchard test results showed that the system accomplished target-oriented fertilization 95 times for 92 trees at the mature stage. The results indicated that the dual-model detection method could be used for Wolfberry trees or other trees with similar canopy changes at different growth stages. Keywords: orchard fertilizer, target-oriented fertilization, variable rate fertilization, photoelectric sensor, wolfberry orchard DOI: 10.25165/j.ijabe.20181104.3614 Citation: Yang S, Zhai C Y, Long J, Zhang B, Li H Z. Wolfberry tree dual-model detection method and orchard target-oriented fertilization system based on photoelectric sensors. Int J Agric & Biol Eng, 2018; 11(4): 65-73.

<abstract> <bold>Abstract.</bold> Generally growers apply fertilizer uniformly once in two year production cycle in wild blueberry fields without considering substantial variation in soil/plant characteristics, topographic features and fruit yield within wild blueberry fields. Split site-specific fertilization may increase crop yield and reduce ammonia volatilization losses. A commercial wild blueberry field was selected in Nova Scotia to evaluate the impact of variable rate split (VRS) fertilization on ammonia volatilization losses. Management zones (MZs) i.e. steep slope (Zone-1), moderate slope (Zone-2) and low lying areas (Zone-3) were delineated based on slope variation within the field. The field was divided into two sections (VRS section and Uniform rate (UR) section). For fertilization GPS-guided prescription map was generated in ArcGIS 10 software. In VRS section, fertilizer was applied three times in Zone-1, Zone-2, and Zone-3 with fertilization rates of 66, 50, and 33 kg ha-1, respectively, using a variable rate (VR) fertilizer spreader. For comparison, single uniform fertilizer application at rate of 200 kg ha-1 was performed in UR section of the selected field. After fertilization, ammonia volatilization trials were established under VRS and UR fertilization in the field. The ammonia volatilization samples were collected after each fertilization, and samples were analyzed for ammonium nitrogen (NH4+-N) concentrations. Leaf samples were taken in third week of July at tip die-back stage and analyzed for macro and micro nutrients. The results of this study suggested that ammonia volatilization losses in VRS section were significantly lower than UR section. It is also observed that the VRS fertilization did not influence negatively to leaf nutrient concentrations. Split fertilizer application in wild blueberry fields can increase input use efficiency, farm profitability, and can reduce environmental contamination.

For the past few years, the automation of transportation becomes a hot research topic for smart cities. Intelligent Transportation System (ITS) aims to manage and optimize the traffic congestion, road accidents, parking allocation using Autonomous Vehicles (AV) system, where the AVs are internally connected for message passing and critical decision making in time-sensitive applications. The data security in such applications can be offered using Blockchain (BC) technology. But, as per the existing literature, there exists no system which can call AVs automatically based on the situation, i.e., call an ambulance in case of an accident, call logistic service in case of home transfer, and call the traffic department in case of traffic jam. Motivated from the aforementioned reasons, in this article, we propose a BC-based secure and intelligent sensing and tracking architecture for AV system using beyond 5G communication network. Recently, AVs are facing issues with sensing and tracking technology as well as the data thefts. AV system contains sensitive information and transfers it through a communication channel to Connected AVs (CAVs), where the corrupted information or delay of a fraction of a second can lead to a critical situation. So, here we present possib the attacks and safety countermeasures using BC technology to protect the AV system. The proposed architecture ensures secure sensing and tracking of an object through BC by deploying AI algorithms at the edge servers. Also, the beyond 5G network enables communications with low latency and high reliability to meet the desires of the aforementioned time-sensitive applications. The proposed system is evaluated by considering the parameters as mobility and data transfer time against the traditional LTE-A and 5G communication networks. The proposed system outperforms traditional systems and can be suitable for diverse applications where latency, reliability, and security are the prime concerns.

Transitioning to a driverless city: Evaluating a hybrid system for autonomous and non-autonomous vehicles

According to the world health organization, there were 1.25 million road accident deaths globally in 2013. To decrease this count, it has been said that autonomous vehicles can play a major role. The autonomous vehicle uses different sensors such as LIDAR, RADAR, Ultrasonic sensors, and cameras for detection of lane lines, objects, traffic sign, traffic lights, and driving path. There are several successful projects on autonomous vehicles, but the main limitations are high costs and regulations. The camera can be a cost-effective solution as compare to LIDAR in all the basic detection needed for autonomous driving. The proper placement of sensors on the vehicle plays an important role in efficient design, working and reducing the cost of the overall autonomous vehicle system. This paper discusses the parameters considered while deciding the positioning of the camera, and the geometric calculation for the best positioning of a monocular camera on an autonomous vehicle.

Underwater imaging is being increasingly helpful for the autonomous robots to reconstruct and map the marine environments which is fundamental for searching for pipelines or wreckages in depth waters. In this context, the accuracy of the information obtained from the environment is of extremely importance. This work presents a study about the accuracy of a reconfigurable stereo vision system while determining a dense disparity estimation for underwater imaging. The idea is to explore the advantage of this kind of system for underwater autonomous vehicles (AUV) since varying parameters like the baseline and the pose of the cameras make possible to extract accurate 3D information at different distances between the AUV and the scene. Therefore, the impact of these parameters is analyzed using a metric error of the point cloud acquired by a stereoscopic system. Furthermore, results obtained directly from an underwater environment proved that a reconfigurable stereo system can have some advantages for autonomous vehicles since, in some trials, the error was reduced by 0.05m for distances between 1.125 and 2.675 m.

Current developments in autonomous vehicle systems typically consider solutions to single problems like road detection, road following and object recognition individually. The integration of these individual systems into a single package becomes difficult because they are less compatible. This paper introduces a generic Integrated Neural System for Autonomous Vehicles (INSAV) package solution with processing blocks that are compatible with each other and are also suitable for hardware implementation. The generic INSAV is designed to account for important problems such as road detection, road structure learning, path tracking and obstacle detection. The paper begins the design of the generic INSAV by building its two most important blocks: the Road Structuring and Path Tracking Blocks. The obtained results from implementing the two blocks demonstrate an average of 92% accuracy of segmenting the road from a given image frame and path tracking of straight roads for stable motion and obstacle detection.

In autonomous vehicle systems, understanding the surrounding environment is mandatory for an intelligent vehicle to make every decision of movement on the road. Knowledge about the neighboring environment enables the vehicle to detect moving objects, especially irregular events such as jaywalking, sudden lane change of the vehicle etc. to avoid collision. This local situation awareness mostly depends on the advanced sensors (e.g. camera, LIDAR, RADAR) added to the vehicle. The main focus of this work is to formulate a problem of reconstructing the vehicle environment using point cloud data from the LIDAR and RGB color images from the camera. Based on a widely used point cloud registration tool such as iterated closest point (ICP), an expectation-maximization (EM)-ICP technique has been proposed to automatically mosaic multiple point cloud sets into a larger one. Motion trajectories of the moving objects are analyzed to address the issue of irregularity detection. Another contribution of this work is the utilization of fusion of color information (from RGB color images captured by the camera) with the three-dimensional point cloud data for better representation of the environment. For better understanding of the surrounding environment, histogram of oriented gradient (HOG) based techniques are exploited to detect pedestrians and vehicles.;Using both camera and LIDAR, an autonomous vehicle can gather information and reconstruct the map of the surrounding environment up to a certain distance. Capability of communicating and cooperating among vehicles can improve the automated driving decisions by providing extended and more precise view of the surroundings. In this work, a transmission power control algorithm is studied along with the adaptive content control algorithm to achieve a more accurate map of the vehicle environment. To exchange the local sensor data among the vehicles, an adaptive communication scheme is proposed that controls the lengths and the contents of the messages depending on the load of the communication channel. The exchange of this information can extend the tracking region of a vehicle beyond the area sensed by its own sensors. In this experiment, a combined effect of power control, and message length and content control algorithm is exploited to improve the map's accuracy of the surroundings in a cooperative automated vehicle system.