Search

US cotton producers are motivated to optimize planter performance to ensure timely and uniform stand establishment early in the season, especially when planting in sub-optimal field conditions. Field studies were conducted in 2017, 2018 and 2019 to evaluate the effect of seeding depth and planter downforce on crop emergence and yield in cotton planted in different soil moisture conditions. Field conditions representative of dry, normal and wet soil moisture conditions were attained by applying 0, 1.27 and 2.54 cm of irrigation within the same field. Two cotton cultivars (representing a small-seeded and a large-seeded cultivar, 9259–10,582 and 11,244–14,330 seeds kg−1, respectively), were planted at seeding depths of 1.3, 2.5 and 3.8 cm with each seeding depth paired with three different planter downforces of 0, 445 and 890 N in each block. Cotton was planted in plots that measured 3.66 m (four-rows) wide by 10.67 m long. Results indicated that crop emergence was affected by the seeding depth across most field conditions and higher crop emergence was observed in the large-seeded cultivar at 1.3 and 3.8 cm seeding depths in dry and wet field conditions, respectively. Lint yield was also higher for the large-seeded cultivar at the 3.8 cm seeding depth across all field conditions in 2017, and in dry field conditions in 2018. Planter downforce effect on crop emergence varied among the cultivars where the large-seeded cultivar exhibited higher crop emergence than the small-seeded cultivar at 445 and 890 N downforce. Planter downforce of 445 N yielded greater than the 0 and 890 N treatment in dry field conditions in 2017. The study results suggest that matching planter depth and downforce settings for prevalent soil moisture conditions at planting along with appropriate cultivar selection can help in achieving optimal emergence and yield in cotton.

Using artificial intelligence (AI) and the IoT (Internet of Things) is a primary focus of applied engineering research to improve agricultural efficiency. This review paper summarizes the engagement of artificial intelligence models and IoT techniques in detecting, classifying, and counting cotton insect pests and corresponding beneficial insects. The effectiveness and limitations of AI and IoT techniques in various cotton agricultural settings were comprehensively reviewed. This review indicates that insects can be detected with an accuracy of between 70 and 98% using camera/microphone sensors and enhanced deep learning algorithms. However, despite the numerous pests and beneficial insects, only a few species were targeted for detection and classification by AI and IoT systems. Not surprisingly, due to the challenges of identifying immature and predatory insects, few studies have designed systems to detect and characterize them. The location of the insects, sufficient data size, concentrated insects on the image, and similarity in species appearance are major obstacles when implementing AI. Similarly, IoT is constrained by a lack of effective field distance between sensors when targeting insects according to their estimated population size. Based on this study, the number of pest species monitored by AI and IoT technologies should be increased while improving the system’s detection accuracy.

HighlightsAn ensemble method using color segmentation, deep learning, and image transformation was developed.Experiments were conducted to compare the method with other state-of-the-art tracking algorithms.The optimized ensemble method to track bolls achieved 94.4% accuracy using weakly trained tiny YOLOv2 models.The method achieved 7.6 frames per second and outperformed five other tracking methods.Abstract. In robotic applications, good perception can be computationally costly and create undesirable latency before a control decision is initiated. Most of the methods available for object detection deep learning are either fast with low accuracy or slow with high accuracy. Fast and accurate methods are necessary to track and localize objects such as cotton bolls that may be visible or occluded by each other or not well illuminated. In this study, an ensemble of a deep learning method and other image processing techniques was used to detect cotton bolls in-field on defoliated plants. In each image, a trained deep learning method, the YOLOv2 model, was used to detect open cotton bolls, and color segmentation was applied to confirm if the bolls detected by the YOLOv2 model were actually white to avoid false positives. Boll tracking was performed by following the spatial movement of good features on the edges of the bolls using the Lucas-Kanade algorithm. An image transformation algorithm was applied to the next image in case the previously detected boll was lost to retrieve the information of the missing boll. Each tracked and localized boll was stored and counted to give the total number of bolls detected. In this study, detection accuracy was sacrificed for image processing speed by using the YOLOv2 model. Detection accuracy was improved by using an ensemble method that combined image color segmentation, optical flow, and image transformation. This method was compared to eight other open-source methods implemented in OpenCV. The ensemble method detected and counted bolls at a speed of 7.6 fps with an accuracy of 94.4% using the Jetson TX2 embedded system to process 1K resolution images, outperforming the other OpenCV methods in various measurements. Keywords: Boll counting, Cotton, Cotton harvesting, DarkFlow, Darknet, Deep learning, Machine vision, YOLOv2.

Maintaining adequate planter downforce is critical for achieving timely and uniform crop emergence across the field. On-farm research studies were conducted in different regions across South Georgia, USA to investigate the effect of planter downforce in varying soil textures. Six fields were planted in cotton with four different growers from 2017 to 2019. Soil electrical conductivity (EC) was mapped in each field to characterize areas of varying soil textures into low, medium and high soil EC zones. For each field, the downforce treatments consisted of a grower-preferred downforce along with a relatively lower and higher downforce dependent on prevalent field conditions and the downforce system available on the grower’s planter. Results varied from field to field depending on soil texture variability or the type of downforce system. In three out of six fields, emergence was delayed and reduced in areas of high soil EC (heavy textured soils) whereas soil EC or downforce did not affect crop emergence in two fields. In one field, increasing planter downforce from the grower preferred value of 555 N to 1110 N improved crop emergence in heavy textured soils. In some cases, it was noted that even the highest downforce (1110 N) was not sufficient to place seeds at the desired seeding depth in the heavy textured soils. More on-farm studies are required to thoroughly understand planter downforce response in varying field conditions that exist during planting. Future studies should consider better quantification of in-field soil variability and changes in planter downforce during field operation.

Peach scab, caused by Venturia carpophila, is a damaging disease of peach in the southeastern United States. Thus, fungicides are applied to reduce peach scab. Tractor speed was investigated as a variable affecting spray deposition and disease control in relation to volume applied. In experiments in 2015 and 2016, trees were sprayed with fungicide to control scab at petal fall to 1% shuck split and at shuck split to 10% shuck off. Speeds were 3.2, 4.8, and 6.4 kph resulting in 1,403, 935, and 701 liters/ha, respectively, with the dose of active ingredient (a.i.) per ha kept constant. Deposition declined for all speeds with later spray dates. There was a negative linear relationship between tractor speed and spray coverage on three of four dates the experiment was repeated. Tractor speed (different volumes, equal doses) affected peach scab. In 2015 and 2016, mean incidence at 3.2, 4.8, and 6.4 kph was 68.6, 59.2, and 38.3%, and 64.2, 53.0, and 40.4% of fruit scabbed, respectively. Effect of speed on lesion number per fruit depended on year: in 2015, lesions per fruit were reduced at 6.4 kph compared with 3.2 and 4.8 kph but were not different in 2016. Control trees had fewer lesions per fruit high in the canopy, but there was little effect of sample height in fungicide-treated trees. Concentration of a.i. in lower volumes applied at higher speed may provide some benefit in reducing incidence of peach scab, but there appeared to be less effect on severity.

Cotton (Gossypium hirsutum L.) growers are motivated to reduce seeding rates due to increased technology fees associated with improved transgenic cotton cultivars. Advances in planting machinery have improved precision of seed metering and seed placement in recent years. A two-year study was conducted to evaluate the effect of seeding rate, planter downforce, and cultivar on crop emergence and lint yield in cotton planted as singulated and hill-drop (two seed hill-1) configuration. Study treatments consisted of two seeding rates (71,660 and 107,490 seed ha-1), two to three planter downforces (0, 445 and 890 N in 2017; 0 and 890 N in 2018) and two cotton cultivars (representing a large-seeded and small-seeded cultivar, 9,259 - 10,582 and 11,244 - 14,330 seed kg-1, respectively) arranged in a strip-split plot design in both seeding configurations. Crop emergence and lint yield in the middle two rows (four-row plots) were measured to evaluate treatment effects among seeding configurations. Results showed that seeding rate and cultivar did not affect (p>0.05) crop emergence and lint yield in both singulated and hill-drop cotton. Crop emergence varied between the two years due to differences in field tillage conditions. Planter downforce affected crop emergence in singulated cotton but not in hill-drop cotton during both years. Field tillage conditions also influenced downforce effect on crop emergence. Selection of an optimal planter downforce had more significant effect (p

Herbicides have been the primary weed management practice in agriculture for decades. However, due to their effects on the environment in addition to weeds becoming resistant, alternative approaches to weed control are critical. One approach is using lasers, particularly diode lasers because of their portability, low power demand, and cost effectiveness. In this research, weeds’ response to diode laser treatments was investigated. Three experiments were conducted. The first experiment involved treating two species of weeds with four different laser powers to determine the time it takes to sever the weed stem. The second experiment involved monitoring the status of two species of weeds for a week after treating them with two lasers at constant application times of 1 s, 2 s, and 3 s. The third experiment was a repeat of the second with higher laser powers and shorter treatment times. The results showed diode lasers have a potential to be an effective weed controlling tool. Weed stem diameter, laser power, treatment duration, and distance between laser and weed were all statistically significant in weed mortality, with weed species having no significance. Furthermore, it was found that weed management is possible by exposing the stem of the two weed species between 0.8 and 2.65 mm diameter to a laser beam dosage without necessarily severing it, with 80% effectiveness at 0.5 s treatment time, and 100% effectiveness using a 6.1 W laser for 1.5 s.

This paper discusses a technique for identifying stress in plants using 4D models to determine stressed areas based on height and growth rates. The developed algorithm segments the field into low, medium, and high stress regions to provide warnings to the grower that an issue might exist. This algorithm was tested on a field of peanut plants and showed that high stressed regions produced 20% lower yield than low stressed regions on average. This metric can provide the grower with an early warning that regions of relatively high stress exists within their field that will have an impact on yield.

Small rovers are being developed to pick cotton as bolls open. The concept is to have several of these rovers move between rows of cotton, and when bolls are detected, use a manipulator to pick the bolls. To accomplish this goal, each cotton-picking robot needs to accomplish three movements; rover must move forward/backward, left/right and the manipulator must be able to move to harvest the detected cotton bolls. Control of these actions can have several states and transitions. Transitions from one state to another can be complex but using ROS-independent finite state machine (SMACH), adaptive and optimal control can be achieved. SMACH provides task level capability to deploy multiple tasks to the rover and manipulator. In this research, a cotton-picking robot using a stereo camera to locate end-effector and cotton bolls is developed. The robot harvests the bolls using a 2D manipulator that moves linearly horizontally and vertically. The boll 3-D position is determined by calculating stereo camera parameters, and the decision of the finite state machine guides the manipulator and the rover to the destination. PID control is deployed to control rover movement to the boll. We demonstrate preliminary results in a direct-sun simulated environment. The system achieved a picking performance of 17.3 seconds per boll. Also, it covered the task by navigating at a speed of 0.87 cm per second collecting 0.06 bolls per second. In each mission, the system was able to detect all the bolls but one.

Here, we present a novel presymptomatic disease detection method by utilizing electrical capacitance tomography (ECT) for plant root monitoring. This was achieved by enhancing the signal with iron oxide nanoparticles (IONPs). IONPs were added to healthy and underdeveloped tomato plants during the course of one and a half months while ECT measurements were collected. The readings for the IONP-exposed samples showed an enhancement over the unexposed samples. This enhancement was greater in the healthy plants when compared with the underdeveloped plants. We also verified earlier claims that ECT can be used for presymptomatic disease detection in bell pepper plants inoculated with Phytophthora capsici ( PCap ). The healthy plants all showed a continuous increase in capacitance over a 2-week period, while the inoculated plants showed a sudden spike in capacitance 12-24 hours before wilting and browning of the stem occurred. The results from these studies indicate that IONP-enhanced electrical capacitance tomography is a non-destructive, in-situ , and rapid method to monitor relative root health, and that ECT can detect disease presymptomatically.

Robotic harvesting involves navigation and environmental perception as first operations before harvesting of the bolls can commence. Navigation is the distance required for a harvester‘s arm to reach the cotton boll while perception is the position of the boll relative to surrounding environment. These two operations give a 3D position of the cotton boll for picking and can only be achieved by detection and tracking of the cotton bolls in real-time. It means detection, tracking and counting of cotton bolls using a moving camera allows the robotic machine to harvest easily. GPU-accelerated deep neural networks were used to train the convolution networks for detection of cotton bolls. It was achieved by using pretrained tiny yolo weights and DarkFlow, a framework which translates YOLOv2 darknet neural networks to TensorFlow. A method to connect tracklets using vectors that are predicted using Lucas-Kanade algorithm and optimized using robust L-estimators and homography transformation is proposed. The system was tested in defoliated cotton plants during the spring of 2018. Using three video treatments, the counting performance accuracy was around 93% with standard deviation 6%. The system average processing speed was 21 fps in desktop computer and 3.9 fps in embedded system. Detection of the system achieved an accuracy and sensitivity of 93% while precision was 99.9% and F1 score was 1. The Tukey‘s test showed that the system accuracy and sensitivity was the same when the plants were rearranged. This performance is crucial for real-time robot decisions that also measure yield while harvesting.

Helicoverpa zea (Boddie) is a major insect pest of corn and other agricultural crops. An improved understanding of semiochemcials that control adult behavior is needed to develop alternative control measures. In this study, overnight SPME collection of volatiles from corn ears enclosed in Teflon bags in the field at two stages of development were made. C8–C10 aldehydes, a C8-alcohol, C6–C9 alcohol acetates, and numerous monoterpenes, sesquiterpenes, sequiterpene alcohols, and geosmin were identified after thermal desorption and GC/MS. Structural assignments of the alcohol acetates, Z-3-hexenyl acetate, 2-heptyl acetate, 2-nonyl acetate, and 4-nonyl acetate, the monoterpenes, α- and β-ocimene, and geosmin were made by analysis of standards that were purchased or prepared in the laboratory. All other assignments were based on published Kovat’s retention time indices (KI) and mass spectra. Pair-wise comparison of the relative amounts of each component between two groups of corn ears defined by silk weight did not identify significant differences, thus it is unknown whether or not silk weight impacted volatile emission composition and rate. To our knowledge three compounds detected in SPME collections, 2-heptyl acetate, 2-nonyl acetate, and 4-nonyl acetate have not been previously reported in corn ear or silk volatiles. Their impact on the flight response of gravid earworm females was evaluated in a flight chamber. No significant response to the individual compounds or a blend of all three was observed. Thus, their impact on moth behavior remains uncertain.

Damage to seedling cotton as a result of feeding by adult and immature thrips can result in stunted growth and reduced plant stand. Transmission of Tomato spotted wilt virus by thrips in peanut can result in reduced yield and plant stand. The objectives of this research were to examine the effects of planting cotton and peanut into conventionally (no soil residue) tilled plots or planting into reduced tillage plots including rolled rye and standing rye (cotton only) tillage in conjunction with thiamethoxam seed treatments on seedling thrips counts, thrips associated injury in cotton, Tomato spotted wilt virus incidence in peanut, and yield for both crops. Conservation tillage practices consistently reduced adult and immature thrips populations in cotton. Standing rye residues suppressed thrips similarly to rolled rye, but yielded less lint in one of the two years. Thiamethoxam seed treatment effects were inconsistent, although general numerical trends suggest that this practice can suppress immature thrips for up to three weeks in cotton. Thiamethoxam seed treatment either had no effect or significantly improved cotton lint yield depending on year. Conservation tillage was also effective at reducing immature thrips in peanut. Incidence of Tomato spotted wilt virus in peanut was decreased in conservation tilled plots in 2013, but was slightly greater in 2014. Similarly, effects of tillage on peanut yield were significant but inconsistent between years. Thiamethoxam seed treatments did not consistently reduce thrips in peanut nor did it improve yield in either year. These data demonstrate conservation tillage and thiamethoxam seed treatments are generally effective practices for reducing thrips densities; however, effects on yield and disease transmission were inconsistent.

The overall goal of this study was to test a customized gas sensor array in its ability to detect an important postharvest disease (sour skin) in onions. The sensor array consists of seven metal oxide semiconductor gas sensors and a microcontroller-based automatic data logging system. Three features were extracted from the sensor responses and three baseline correction methods were employed to correct the sensors’ responses. The gas sensor array was tested in two separate experiments with two treatments (control and sour skin). The multivariate data analysis revealed that the “relative response” feature combined with relative baseline correction method provided the best discrimination of infected onions among healthy ones. The best performance (85%) was achieved by using the support vector machine model when the data collected from an independent experiment were used for validation. The study demonstrated the potential of a gas sensor array to detect sour skin-infected onions placed among healthy onions in storage.

A gas sensor array, consisting of seven Metal Oxide Semiconductor (MOS) sensors that are sensitive to a wide range of organic volatile compounds was developed to detect rotten onions during storage. These MOS sensors were enclosed in a specially designed Teflon chamber equipped with a gas delivery system to pump volatiles from the onion samples into the chamber. The electronic circuit mainly comprised a microcontroller, non-volatile memory chip, and trickle-charge real time clock chip, serial communication chip, and parallel LCD panel. User preferences are communicated with the on-board microcontroller through a graphical user interface developed using LabVIEW. The developed gas sensor array was characterized and the discrimination potential was tested by exposing it to three different concentrations of acetone (ketone), acetonitrile (nitrile), ethyl acetate (ester), and ethanol (alcohol). The gas sensor array could differentiate the four chemicals of same concentrations and different concentrations within the chemical with significant difference. Experiment results also showed that the system was able to discriminate two concentrations (196 and 1964 ppm) of methlypropyl sulfide and two concentrations (145 and 1452 ppm) of 2-nonanone, two key volatile compounds emitted by rotten onions. As a proof of concept, the gas sensor array was able to achieve 89% correct classification of sour skin infected onions. The customized low-cost gas sensor array could be a useful tool to detect onion postharvest diseases in storage.

Autonomous crop monitoring at high spatial and temporal resolution is a critical problem in precision agriculture. While Structure from Motion and Multi-View Stereo algorithms can finely reconstruct the 3D structure of a field with low-cost image sensors, these algorithms fail to capture the dynamic nature of continuously growing crops. In this paper we propose a 4D reconstruction approach to crop monitoring, which employs a spatio-temporal model of dynamic scenes that is useful for precision agriculture applications. Additionally, we provide a robust data association algorithm to address the problem of large appearance changes due to scenes being viewed from different angles at different points in time, which is critical to achieving 4D reconstruction. Finally, we collected a high quality dataset with ground truth statistics to evaluate the performance of our method. We demonstrate that our 4D reconstruction approach provides models that are qualitatively correct with respect to visual appearance and quantitatively accurate when measured against the ground truth geometric properties of the monitored crops., Comment: Submitted to IEEE International Conference on Robotics and Automation (ICRA) 2017

A pure-pursuit navigational control algorithm in an autonomous articulated-steer vehicle was studied as a simple method for path-following. Paths were created using GPS coordinates and the articulated-steer vehicle programmed to follow those paths using a pure-pursuit (goal-seeking) path-following algorithm. A GUI was developed in LabVIEW to read RTK-GPS coordinates, hydraulic pressure and articulation angle, and control the articulated-steer vehicle. The vehicle was tested for its ability to follow five different paths with two speeds and three control pulse signals. Path following succeeded with lower vehicle speed and full pulse signal for the straight and sinusoidal paths. Averaged path-following errors ranged from 6 to 19 cm as the turn radius increased. For a 90 degree turn, the path-following error was almost 80 cm. Path-following improvements could be made with faster control signal updates and a variable look-ahead distance based on vehicle speed and required turning angle. It was concluded that the control architecture is a good basis for navigation of an articulated-steer autonomous vehicle.

PURPOSE: This study used a randomized control design to evaluate the effectiveness of AgTeen, an in-home, family-based farm safety intervention, in decreasing extra riding on tractors by youth. Having children as extra riders on tractors has deep roots in farm culture, but it can result in serious injury or death. METHODS: The study randomized 151 families into 3 groups: parent-led intervention (fathers taught their families about farm safety), staff-led intervention (staff members who were peer farmers taught families), and a no-treatment control. Mothers, fathers, and all children aged 10-19 participated in the lessons. FINDINGS: At study entry, 93% of youth reported that they had been an extra rider on a tractor in the past year. Although they were aware of the injury risk, fathers frequently gave tractor rides to their children. After the intervention, fathers in both AgTeen groups were less likely than control fathers to give youth tractor rides. Intervention youth were less likely than control youth to be extra riders. The intervention positively affected the extra-riding attitudes and injury risk perceptions of mothers and fathers. The parent-led and staff-led groups did not significantly differ across study outcomes. CONCLUSIONS: Findings confirm the effectiveness of a family-based intervention in decreasing extra riding on tractors by youth. Language: en

PURPOSE: Farm youth continue to experience high rates of injury and deaths as a result of agricultural activities. Farm machinery, especially tractors, is the most common cause of casualties to youth. A Roll-Over Protection Structure (ROPS) along with a fastened seatbelt can prevent almost all injuries and fatalities from tractor overturns. Despite this knowledge, the use of seatbelts by farmers on ROPS tractors remains low. This study treats farm safety as a family issue and builds on the central role of parents as teachers and role models of farm safety for youth. METHODS: This research study used a longitudinal, repeated-measures, randomized-control design in which youth 10-19 years of age were randomly assigned to either of two intervention groups (parent-led group and staff-led group) or the control group. RESULTS: Fathers in the parent-led group were less likely to operate ROPS tractors without a seatbelt compared with other groups. They were more likely to have communicated with youth about the importance of wearing seatbelts on ROPS tractors. Consequently, youth in the parent-led group were less likely to operate a ROPS tractor without a seatbelt than the control group at post-test. CONCLUSIONS: This randomized control trial supports the effectiveness of a home-based, father-led farm safety intervention as a promising strategy for reducing youth as well as father-unsafe behaviors (related to tractor seatbelts) on the farm. This intervention appealed to fathers' strong motivation to practice tractor safety for the sake of their youth. Involving fathers helped change both father as well as youth unsafe tractor-seatbelt behaviors. Language: en

Blueberry mechanical harvesters cause bruise damage to the fruit. The goal of this study was to test a custom-made sensor (berry impact recording device) to measure the quality and magnitude of mechanical impacts created by three major types of commercial blueberry mechanical harvesters (rotary, slapper, and sway). The sensor was mounted on blueberry bushes (cultivated) and harvested at standard operating conditions such that the sensor was detached and experienced the impact forces typically found during a mechanical harvesting process. The data collected by the sensor revealed that the slapper and sway harvesters generated not only larger number but also higher magnitude impacts than the rotary. Our analyses suggest that these disparities were mostly caused by different agitating mechanisms, contacting surface materials, and designs between the three harvesters. Results indicated that most impacts lasted 5-7ms in all three harvesters. The distribution of the impacts showed that 90% of impacts from the rotary were less than 190g and 90% of impacts from the slapper and sway were less than 250g. Corresponding measures were identified to reduce potential bruise damage in the harvesters. The information could be useful to select harvesters that create the least impacts and to improve current mechanical harvester designs.

The objective of this research was to test the applicability of using electrical capacitance tomography (ECT) as a rapid, in situ, and non-destructive method to monitor root health and detect disease before visual symptoms present. Two studies were performed. The first study compared capacitance measurements of healthy and stunted tomato plants. ECT measurements showed a slight increase (34%) in capacitance for the stunted group, while the healthy plants displayed a large increase (67%). The second study compared ECT measurements of healthy bell pepper plants and pepper plants exposed to Phytophthora capsici (PCap). The inoculated plants that died showed a sudden spike in capacitance (~52%) 12-24 hours before wilting and browning of the stem occurred. The healthy plants all showed a greater increase in capacitance compared with inoculated plants that survived for the same monitoring period. The results from the two studies indicate that ECT is a rapid and non-destructive method to monitor root health and detect disease pre-symptomatically.

Southern highbush blueberries (Vaccinium corymbosum interspecific hybrids) are predominantly for the fresh market and, with a few exceptions, are hand harvested. Though mechanical harvesting systems are available for processed blueberries, low harvest efficiency and high fruit damage have limited their use for picking blueberries for fresh market. To improve the current machine harvester design, it is apparent that the interaction between the machine harvester and fruit should be well understood. The goal of this study was, therefore, to provide such an understanding by using a custom-made miniature instrumented sphere. The miniature Berry Impact Recording Device (BIRD) was used to measure the mechanical impacts created by a rotary mechanical harvester during fruit harvesting. A closeup video recorded the harvesting to pinpoint critical control points where most impacts were created. The results showed that the catch plates on the rotary harvester accounted for over 30% of all mechanical impacts imposed on the BIRD, followed by the empty fruit collection box (lug) (>20%). Impacts created by the conveyer belt and shaking rods combined accounted for only 25% of mechanical impacts. Thus, the most significant reduction in bruising could be achieved through modifications in the catch plates and lugs. The impact of three contacting surfaces (the catch plates, conveyer belt, and steel tunnel) on the harvester and three commercial padding materials (Cellular Silicone, Slow Rebound, and No Bruze) were evaluated. Harvester surface evaluations revealed that the catch plate was the hardest surface. Among the three padding materials evaluated, the Cellular Silicone provided the best cushioning. This study proved the efficacy of using a customized miniature instrumented sphere in measuring mechanical impacts created by various machine parts during harvesting. It provided a better understanding of how the berries interact with different machine parts of a rotary harvester.

Developing data acquisition software is a major challenge in integrating a spectral imaging system. This paper presents the design and implementation of a data acquisition program using LabVIEW for a liquid crystal tunable filter based spectral imaging system (900-1700nm). The module-based program was designed in a three-tier structure. The image acquisition process, modelled by a finite state machine, was implemented in LabVIEW to control the spectral imaging system to collect hyperspectral or multispectral images. The collected spectral images were encoded in general format and could be further processed by other common spectral image analysis tools. In addition, the program could be used to observe band ratio images of the test object in real-time, collect spectral images after ensemble averaging, and select region of interest for spectral image acquisitions. This program is a useful data acquisition tool for the filter-based spectral imaging system. The design and implementation techniques described in this article could also be used to develop similar spectral image acquisition programs.

Sour skin (Burkholderia cepacia) is a major postharvest disease for onions and causes substantial production and economic losses in onion postharvest. In this study, a shortwave infrared hyperspectral imaging system was explored to detect sour skin. The hyperspectral reflectance images (950–1650 nm) of onions were obtained for the healthy and sour skin-infected onions. Principal component analysis conducted on the spectra of the healthy and sour skin-infected onions suggested that the neck area of the onion at two wavelengths (1070 and 1400 nm) was most indicative of the sour skin. Log-ratio images utilizing the two optimal wavelengths were used for two different image analysis approaches. The first method applied a global threshold (0.45) to segregate the sour skin-infected areas from log-ratio images. Using the pixel number of the segregated areas, Fisher’s discriminant analysis recognized 80% healthy and sour skin-infected onions. The second classification approach used three parameters (max, contrast, and homogeneity) of the log-ratio images as the input features of support vector machine (Gaussian kernel, γ = 1.5), which discriminated 87.14% healthy and sour skin-infected onions. The result of this study can be used to further develop a multispectral imaging system to detect sour skin-infected onions on packing lines.

Calibration is a critical step for developing spectral imaging systems. This paper presents a systematic calibration and characterization approach for a liquid crystal tunable filter (LCTF) based shortwave infrared (SWIR) spectral imaging system. A series of tests were conducted to validate the linearity of the system output, measure the field of view of the spectral imager, increase the system spectral sensitivity, test the spatial and spectral resolution of the system, evaluate the system stability and image distortion, and reduce the spectral noise of the system output. Results showed that the system had an angle of view of 6.98^o and a spatial resolution of 158@mm. The spectral sensitivity of the system was corrected by controlling the camera exposure time and gain, which increased the signal to noise ratio of the system by 16.5%. Test results also verified the system spectral accuracy and linearity (r>0.999). The system output was proven to be stable and image distortion was not perceivable. Results of calibration tests indicated that this system satisfied the design criteria in both spatial and spectral domains. The calibration methods presented here are applicable to the LCTF-based spectral imaging systems in other applications.

This paper presents the methodology to design and integrate a liquid crystal tunable filter (LCTF) based shortwave infrared (SWIR) spectral imaging system. The system consisted of an LCTF-based SWIR spectral imager, an illumination unit, a frame grabber, and a computer with the data acquisition software. The spectral imager included an InGaAs camera (320x256 pixels), an SWIR lens (50mm, F/1.4), and an LCTF (20mm aperture). Four multifaceted reflector halogen lamps (35W, 12 VDC) were used to build the illumination unit. The system was integrated by a LabVIEW program for data acquisition. It can capture hyperspectral or multispectral images of the test object in the spectral range of 900-1700nm. The system was validated by differentiating sugar from wheat flour, and water from 95% ethanol. The results showed that the system can distinguish these materials in both spectral and spatial domains. This SWIR spectral imaging system could be a potential useful tool for nondestructive inspection of food quality and safety.

This paper reports a complete impact data acquisition, processing, and analyzing software system that applies on the hardware platform of the Berry Impact Recording Device (BIRD). The software has three major sections that correspond to the hardware: The BIRD sensor program, the interface box program, and the computer software i-BIRD. The sensor program samples acceleration data from three axes and records them as single impacts with a maximum sampling rate of 3.0 kHz. Users can configure the sensor via the i-BIRD computer software, with different options of sampling frequencies (682–3050 Hz) and thresholds (0–205 g , where g is the gravitational acceleration). The data recorded can be downloaded, processed and graphically displayed on the computer. A real time clock was created using the interrupt service routine provided by the microcontroller. The accuracy of the sensor’s clock was calibrated with an error of 0.073%, which was adequate to record impact data in this application. The shape of impact curves recorded by the BIRD sensor at three sampling frequencies (682, 998, and 1480 Hz) matched well with the curves recorded by a high frequency (10 kHz) data logger with the maximum root mean squared error of 4.4 g. The velocity change had a relative error less than 5%. With confirmation of all those performances, the software system enabled the BIRD to be a useful tool to collect impact data during small fruit (such as blueberry) mechanical harvest.

Agricultural technology has increased farm production to unprecedented levels. However, return on investment is diminishing and environmental concerns conflict with current input intensive farm practices. Conventional technologies and their application such as crop breeding and management practices have focused on monocultural systems that are dependent on chemical inputs to produce optimum yields. Current profit margins are low or non-existent with these conventional non-sustainable practices and must be changed if the family farm is to survive. We propose an ecologically based approach to farm management that strives to reduce reliance on chemically intensive inputs through better use of multiple attributes inherent within agroecosystems. This approach requires a redirection in the development and application of current and emerging technologies. Examples of redirections in research and development programs for pest management practices, genetic engineering, and precision agriculture necessary to provide a more ecologically-based and sustainable farming approach are illustrated.

The off-flavor boar taint associated with the substances skatole, androstenone, and possibly indole represents a significant problem in the pig husbandry industry. Boar taint may occur in meat from uncastrated sexually mature male pigs; consumers commonly show a strong aversion to tainted meat. Consequently, there is a need for rapid methods to sort out and remove tainted carcasses at the slaughterline. We tested the ability of wasps, Microplitis croceipes to perceive and learn the 3 boar taint compounds both individually and in combination using classical conditioning paradigms. We also established the effectiveness and reliability of boar taint odor detection when wasps were used as biosensors in a contained system called the “wasp hound” using a cohort of trained wasps. We found that the wasps are able to successfully learn indole, skatole and to also detect them when presented a 1: 1: 1 mixture of all 3 compounds. This was shown for both a single hand-manipulated wasp bioassay and when using the “wasp hound” detector device. In contrast, the wasps showed a weak conditioned response to androstenone at the concentration tested. The estimated gas phase concentrations that the wasps perceived during training were in the range of 10 ± 0.4 pg/s for skatole and indole, and 2 ± 0.5 pg/s for androstenone. We conclude that use of these wasps as biosensors presents a promising method for boar taint detection and discuss future training paradigms that may improve their responses to compounds such as androstenone. Practical Application: The development of a perceptive, inexpensive, and reliable means of detecting boar taint before the product is presented to sensitive consumers.

Palmer amaranth is resistant to several herbicides, including glyphosate, and there is concern that the resistance traits can be transferred between spatially segregated populations via pollen movement. The objective of this study was to describe the physical properties of Palmer amaranth pollen, specifically size, density, and settling velocity (Vs), that influence pollen flight. The mean diameter for Palmer amaranth pollen, as determined by light microscopy, was 31 µm (range of 21 to 38 µm); mean pollen diameter as measured with the use of an electronic particle sizer was 27 µm (range of 21 to 35 µm). The mean density of the solid portion of the pollen grain was 1,435 kg m−3. Accounting for the density of the aqueous fraction, the mean density of a fully hydrated pollen grain was 1,218 kg m−3. By Stokes's law, the estimated mean theoreticalVsfor individual Palmer amaranth pollen grains was 3.4 cm s−1for the range of pollen diameters with a mean of 31 µm and 2.6 cm s−1for the range of pollen diameters with a mean of 27 µm. Results from laboratory studies indicated the majority of single pollen grains settled at a rate of 5.0 cm s−1. The difference between the theoretical and empirical estimates ofVswas likely due to changes in pollen density and shape postanthesis, which are not accounted for using Stokes's law, as well as the presence pollen clusters.

Most studies on plant-herbivore interactions focus on either root or shoot herbivory in isolation, but above- and belowground herbivores may interact on a shared host plant. Cotton (Gossypium spp.) produces gossypol and a variety of other gossypol-like terpenoids that exhibit toxicity to a wide range of herbivores and pathogens. Cotton plants also can emit herbivore-induced volatile compounds at the site of damage and systemically on all tissues above the site of damage. As these volatile compounds attract natural enemy species of the herbivore, they are thought to represent an indirect plant defense. Our study quantified gossypol and gossypol-like compounds in cotton plants with foliage feeding (Heliocoverpa zea), root feeding (Meloidogyne incognita), or their combination. Cotton plants with these treatments were studied also with respect to induced local and systemic volatile production and the attraction of the parasitic wasp Microplitis croceipes to those plants. We also evaluated whether foliage or root feeding affected foliar nitrogen levels in cotton. After 48 hr of leaf feeding and 5 wk of root feeding, local and systemic induction of volatiles (known to attract parasitoids such as M. croceipes) occurred with herbivore damage to leaves, and it increased in levels when root herbivory was added. Nevertheless, M. croceipes were equally attracted to plants with both leaf and root damage and leaf damage only. In contrast to previous studies in cotton, production of gossypol and gossypol-like compounds was not induced in leaf and root tissue following foliage or root herbivory, or their combination. We conclude that root feeding by M. incognita has little influence on direct and indirect defenses of Gossypium hirsutum against insect herbivory.

Classical conditioning, a form of associative learning, was first described in the vertebrate literature by Pavlov, but has since been documented for a wide variety of insects. Our knowledge of associative learning by insects began with Karl vonFrisch explaining communication among honeybees, Apis mellifera L. (Hymenoptera: Apidae). Since then, the honey bee has provided us with much of what we understand about associative learning in insects and how we relate the theories of learning in vertebrates to insects. Fruit flies, moths, and parasitic wasps are just a few examples of other insects that have been documented with the ability to learn. A novel direction in research on this topic attempts to harness the ability of insects to learn for the development of biological sensors. Parasitic wasps, especially Microplitis croceipes (Cresson) (Hymenoptera: Braconidae), have been conditioned to detect the odors associated with explosives, food toxins, and cadavers. Honeybees and moths have also been associatively conditioned to several volatiles of interest in forensics and national security. In some cases, handheld devices have been developed to harness the insects and observe conditioned behavioral responses to air samples in an attempt to detect target volatiles. Current research on the development of biological sensors with insects is focusing on factors that influence the learning and memory ability of arthropods as well as potential mathematical techniques for improving the interpretation of the behavioral responses to conditioned stimuli. Chemical detection devices using arthropod-based sensing could be used in situations where trained canines cannot be used (such as toxic environments) or are unavailable, electronic devices are too expensive and/or not of sufficient sensitivity, and when conditioning to target chemicals must be done within minutes of detection. The purpose of this article is to provide a brief review of the development of M. croceipes as a model system for exploring associative learning for the development of biological sensors.

Although female parasitic wasps are known to learn to associate odors with hosts and food, the ability of males to learn and detect odors has been neglected. We conducted laboratory experiments to compare the detection ability of learned odors between males and females in the larval parasitoid Microplitis croceipes (Cresson) (Hymenoptera: Braconidae). We first conditioned males and females to associate sucrose water with methyl benzoate, 3-octanone, or cyclohexanone, and then observed their behavior toward various concentrations (40 ng l -1 -4 mg l -1 ) of the trained odors. Conditioned male wasps responded as well as female wasps to various concentrations of the three odors. Response times by wasps to these three odors were not significantly different between sexes. For the three odors, response times of both sexes were longer at the intermediate concentrations (40-400 μg l -1 ) than the higher or lower concentrations. The present study suggests that M. croceipes males can learn and respond to the three chemicals tested as well as the females, and conditioned males are as sensitive to learned odors as conditioned females. By using their sensitive learning and odor-detection capabilities, M. croceipes males could search for food sources as efficiently as females under natural conditions.

Microplitis croceipes (Cresson) (Hymenoptera: Braconidae) are parasitoid wasps capable of being trained to respond to target odors. One such response is known as area-restricted searching, and several wasps exhibiting area-restricted searching within the same area is known as crowding. A computer vision system consisting of a laptop computer, web camera, and software package (Visual Cortex) was assembled to objectively quantify the crowding behavior of M. croceipes. The system was able to measure the crowding behavior of five female M. croceipes hand-trained to detect 3-octanone. Further, the system was able to distinguish between the crowding response of trained M. croceipes exposed to 5.5 and 1.1 mg L-1 of 3-octanone and a control within 20 s.

We have developed the framework for creating an autonomous field scout (AFS) consisting of a ground vehcle for collecting physical samples and an aerial system working that collects images of plants during the growqing season. Images will be processed to creat 3-D structure of plants and then examined over time to look for trends that may indicate early onset plant stress. The ground system (red rover) will then navigate to those locations identified through aerial imaging and experts analysis and collects leaf and/or soil samples as needed. The rover will return to base and the leaf samples examined by an expert and or sent for laboratory analysis to confirm specific pathogens or soil nutrient deficiencies. Ealry detercvtion and more accurate scouting information is a major goal of the system. In the first year of development, we have identified the components and begun assembly of the rover and testing of the robotic arm for leaf and soil sampling. Data will be collected using a tractor, GNSS and a UAS this year until the red rover is ready for operation in year 2. The system will be developed in the ROS, which has integrated OpenCV and Point Cloud Library, for navigation and control.

The ability of many insects to learn has been documented. However, a limited number of studies examining associative learning in medically important arthropods has been published. Investigations into the associative learning capabilities of Culex quinquefasciatus Say were conducted by adapting methods commonly used in experiments involving Hymenoptera. Male and female mosquitoes were able to learn a conditioned stimulus that consisted of an odor not normally encountered in nature (synthetic strawberry or vanilla extracts) in association with an unconditioned stimulus consisting of either a sugar (males and females) or blood (females) meal. Such information could lead to a better understanding of the ability of mosquitoes to locate and select host and food resources in nature.

A portable, handheld volatile odor detector ("Wasp Hound") that utilizes a computer vision system and Microplitis croceipes (Cresson) (Hymenoptera: Braconidae), a parasitoid wasp, as the chemical sensor was created. Five wasps were placed in a test cartridge and placed inside the device. Wasps were either untrained or trained by associative learning to detect 3-octanone, a common fungal volatile chemical. The Wasp Hound sampled air from the headspace of corn samples prepared within the lab and, coupled with Visual Cortex, a software program developed using the LabView graphical programming language, monitored and analyzed wasp behavior. The Wasp Hound, with conditioned wasps, was able to detect 0.5 mg of 3-octanone within a 240 mL glass container filled with feed corn ( approximately 2.6 x 10(-5) mol/L). The Wasp Hound response to the control (corn alone) and a different chemical placed in the corn (0.5 mg of myrcene) was significantly different than the response to the 3-octanone. Wasp Hound results from untrained wasps were significantly different from trained wasps when comparing the responses to 3-octanone. The Wasp Hound may provide a unique method for monitoring grains, peanuts, and tree nuts for fungal growth associated with toxin production, as well as detecting chemicals associated with forensic investigations and plant/animal disease.

The parasitic wasp Microplitis croceipes (Cresson) (Hymenoptera: Braconidae) learns to associate odors with food resources and subsequently exhibits a characteristic food-seeking behavior when encountering the learned odor. Wasps so conditioned, learned and subsequently demonstrated an ability to distinguish among aflatoxigenic and non-aflatoxigenic Aspergillus flavus and A. parasiticus strains. The effects of fungal species, strain, age (5, 10-12, 20, and 30 d) and growth media (potato dextrose agar, peanut agar and corn agar) on the learning and recognition responses of the conditioned wasps were examined. The level of differentiation between fungal strains by conditioned wasps was lowest when working with 5-d-old fungal cultures but increased with age and generally peaked with 20-d-old fungi. Wasps responded generally stronger to the fungal strain conditioned to independent of growth media. This ability of parasitic wasps to learn and distinguish fungal odors can open new avenues in insect learning.

This study was designed to identify and compare the Fusarium species of the Gibberella fujikuroi complex on pearl millet (Pennisetum glaucum (L.) R. Br) and corn (Zea mays L.) crops grown in southern Georgia, and to determine their influence on potential fumonisin production. Pearl millet and corn samples were collected in Georgia in 1996, 1997 and 1998. Three percent of the pearl millet seeds had fungi similar to the Fusarium species of the G. fujikuroi species complex. One hundred and nineteen representative isolates visually similar to the G. fujikuroi species complex from pearl millet were paired with mating population A (Fusarium verticillioides (Sacc.) Nirenberg), mating population D (F. proliferatum (Matsushima) Nirenberg) and mating population F (F. thapsinum (Klittich, Leslie, Nelson and Marasas) tester strains. Successful crosses were obtained with 50.4%, 10.1% and 0.0% of these isolates with the A, D and F tester strains, while 39.5 of the isolates did not form perithecia with any tester strains. Two of the typical infertile isolates were characterized by DNA sequence comparisons and were identified as Fusarium pseudonygamai (Nirenberg and O’Donnell), which is the first known isolation of this species in the United States. Based on the pattern of cross-compatibility, conidiogenesis, colony characteristics and media pigmentation, a majority of the infertile isolates belong to this species. Fumonisins FB 1 and FB 2 were not detected in any of the 81 pearl millet samples analyzed. The species of the G. fujikuroi species complex were dominant in corn and were isolated from 84%, 74% and 65% of the seed in 1996, 1997 and 1998, respectively. Representative species of the G. fujikuroi species complex were isolated from 1996 to 1998 Georgia corn survey (162, 104 and 111 isolates, respectively) and tested for mating compatibility. The incidence of isolates belonging to mating population A (F. verticillioides) ranged from 70.2% to 89.5%. Corn survey samples were assayed for fumonisins, and 63% to 91% of the 1996, 1997 and 1998 samples were contaminated. The total amount of fumonisins in the corn samples ranged from 0.6 to 33.3 μg/g.

An Agleader cotton yield monitor was modified for use as a peanut yield monitor for harvest during the 2000, 2001, and 2002 seasons. A single pair of optical sensors was mounted on the vertical portion of the delivery chute between the air fan and basket. The first season’s harvest resulted in severe pitting and scarring of a glass plate used to protect the sensors from peanut and foreign material abrasion. Even so, mean load absolute error was only 11.6% for 25 loads of peanuts harvested. For the second and third seasons, a deflection baffle and a small air circulation opening at the bottom of each sensor were made to prevent abrasion from peanuts and foreign material and to reduce dust build-up, respectively. Inspection of the sensors showed that the modifications did prevent pitting and scarring on the sensor covering and that the dust build-up was reduced significantly. Absolute mean load error for three different fields harvested was 9.4%, 3.2%, and 4.9%. It was determined that adjustments to the combine at the beginning of the harvest season created the larger error in the first field. Once the machine was adjusted and the yield monitor calibrated to that setting, the Agleader yield monitor results were comparable to the results when used to monitor cotton yield.

We examined the ability of M. croceipes to learn, detect, and respond to 2,4-DNT, which is a volatile discriminator of trinitrotoluene (TNT). The percentage of conditioned wasps to detect and respond to the various concentrations of 2,4-DNT for > or = 15 sec was measured. Significantly more of the conditioned wasps responded to the concentration of 2,4-DNT used for conditioning than other concentrations examined. Accordingly, percent conditioned wasps to respond > or = 15 sec could be used as a suitable measure to screen air samples and distinguish between samples with or without the target odorant. The data recorded in this study indicate the measured behavior could be used to estimate the concentration of target odorants. Data in this study indicate M. croceipes can detect and respond to this compound, which provide further support for its development as a biological sensor.

Based on a novel cotton-flow sensor, the Mississippi Cotton Yield Monitor has been under development at Mississippi State University since 1999, when one prototype of the yield monitor was field tested in Mississippi. Three prototypes were constructed and field tested in Texas, Georgia, and Mississippi in 2000. Five prototypes of an improved version were fabricated and field tested in 2001. All three years’ testing results were promising. In 2001, Mississippi State University and Agricultural Information Management, LLC, signed a licensing option agreement to prepare for the eventual manufacture and marketing of the Mississippi cotton yield monitor. Research towards commercialization of the Mississippi cotton yield monitor was conducted in 2002. In this beta test, 10 prototypes of the Mississippi cotton yield monitor were built and extensively tested on commercial and research farms during the 2002 cotton harvesting season. Harvesting was conducted with both cotton pickers and strippers. The yield monitor system’s accuracy was evaluated on a load-by-load weight basis. Reliability was tested under commercial harvesting conditions. All systems performed well during the tests. Each one was easy to install, maintain, and operate. No hardware problems occurred. The system’s average absolute error was 3.8%. Cotton yield maps created with the data collected by the monitors realistically exhibited yield variations within the fields. Evaluator’s suggestions, mainly related to software performance, are to be addressed in the commercial version of the Mississippi Cotton Yield Monitor. © 2003 Elsevier B.V. All rights reserved.

Microplitis croceipes (Cresson) (Hymenoptera: Braconidae) learns odors in association with both hosts and food. The food-associated ‘seeking’ behavior of M. croceipes was investigated under various training protocols utilizing the conditioning odor, 3-octanone. We investigated the effects of odor training, or its lack, training duration, training frequency, time elapsed after training, wasp hunger state, and training reinforcement, on the food-seeking responses of M. croceipes females. We found that odor-trained females show strong food seeking responses, whereas non-odor-trained females do not respond to the odor, and that a single 10 s association with the odor whilst feeding on sugar water subsequently conditioned the wasps to exhibiting significant responses to it. Increases in training time to more than 10 s did not improve their responses. Repetition of the food‐ odor associations increased a wasp’s recall, as well as its response over time, compared to a single exposure. Repeated exposure to the learned odor in the absence of a food reward decreased the responses of less hungry individuals. However, the level of response increased significantly following a single reinforcement with the food‐ odor association. Understanding the factors that influence learning in parasitoids can enhance our ability to predict their foraging behavior, and opens up avenues for the development of effective biological detectors.

Volatile chemical signals are used by many animals to find food, mates, or hosts. While the keen sense of smell of dogs has been used for centuries, other animals have not been significantly utilized. Recent studies have indicated that many insect species have the ability to learn volatile chemical compounds in association with food or other resources. These insects present a novel approach to volatile chemical detection that could provide a highly sensitive, inexpensive, flexible, and portable sensor. One characteristic of insects that makes them desirable as a potential chemical detector is their ability to detect extremely low levels of chemical compounds. A parasitoid wasp, Microplitis croceipes, was used as the model insect for determining the threshold of response for four compounds: 3-octanone, a compound found in many fungal pathogens; myrcene, a volatile constituent released by cotton plants fed on by cotton bollworms; and putriscene and cadaverine, two products of the breakdown of dead animal protein by microorganisms. Eighteen wasps were trained to each of these individual compounds at one dosage and tested at decreasing dosage levels until their responses were negligible. Each dosage was tested with 18 freshly trained wasps. The wasp response to the odor was determined by a searching behavior called antennating. Wasp response was measured by the length of time the wasp antennated when exposed to the odor. The mean wasp response fell below 10 s at approximately 3.1 × 10-7, 2.9 × 10-7, 3.9 × 10-6, and 4.5 × 10-7 mol L-1 of compound for 3-octanone, myrcene, cadaverine, and putriscene, respectively. For comparative purposes, the detection limits of an electronic nose, the Cyranose 320, was determined for two of the four compounds. The response limits of the wasp for the compounds 3-octanone and myrcene were 74 and 94 times better than the electronic nose, respectively. The response limit of the wasps to putriscene, 3-octanone, and myrcene was approximately 10 times better than to cadaverine.

Establishing a set of electronic multimedia educational programs will expedite the understanding and implementation of key precision agriculture technologies. Several universities and private sector partners have joined together to accomplish this task and to make the material relevant to a diverse audience. Sections include base subject information, grower experiences, potential applications, guidance on how to determine the value of application techniques, and links to additional information. The learning modules (provided in a multimedia format) include information on Global Positioning Systems (GPS), Variable Rate Technologies (VRT), Remote Sensing (RS), sampling and scouting techniques, yield and quality monitoring, and Geographic Information Systems (GIS). The programs of the University of Georgia (UGA) and of Washington State University (WSU) are highlighted. A discussion of the fundamentals of electronic publishing is included. The design of a Internet-era educational program is crucial not only so that the material is conveyed in an efficient and effective way, but also so that ongoing maintenance is minimized.

Researchers at GTRI and UGA Tifton are collaborating on the design, fabrication, and deployment of a low-cost robot arm that is over 80% 3D-printed from plastic. The targeted first application is inspecting bell pepper leaves for the early detection of pests, for which the robot arm will be mounted onto one of UGA’s autonomous rovers that are capable of navigating up and down rows of a crop autonomously. The arm will be able to carry multiple sensors or grappling packages and used in sensing and harvesting tasks. The intent of this particular study is to demonstrate the efficacy of using consumer-level 3D printing technology for an application that requires both precision and repeatability. Due to the particular geometry of the rover and the layout of the bell pepper plantings, the reach of the arm is about 5 feet measured from its anchor point on the rover. The payload for this application consists of a suite of sensors that are approximately 2 pounds. Two challenges result from the design parameters of this application: (1) the long reach of the robot arm, made mostly from relatively flexible plastic, can lead to imprecision and vibrations during motion control and (2) a 2-pound payload cantilevered several feet represents a significant weight to support.

In this study, we propose a novel technique for imaging of crops. Related work has significant limitations in terms of resolution as well as coverage in space and time. In order to overcome these limitations, we use a dense sampling of high-resolution images, which enables the creation of a sequence of 3D reconstructions over time, i.e., a 4D spatio-temporal model of a crop as it develops. To this end, we mounted a CCD camera with integrated GPS and Inertial Measurement Unit (IMU) on a ground articulated tractor, and we collected a set of images approximately once per week over staggered plantings of broccoli from October to January in South Georgia. Multiple images of individual plants taken from different angles are stitched together to produce a 3D reconstruction each week, using “Structure from Motion” (SFM) techniques from computer vision. These 3D reconstructions are then combined in a 4D GIS to show the plants growth and production of fruit over time (4D). These 4D imaging techniques can be applied to images collected from autonomous ground or aerial vehicles for multiple site-specific management practices, such as estimating where, when and how many fruit are ready for harvest in certain vegetable and fruit crops. Other potential applications include record keeping of canopy development for orchards, measuring rate of change of leaf area index, and weed dispersal, proliferation and identification.

A prototype chemical sensor named Wasp hound® that utilizes five classically conditioned parasitoid wasps, Microplitis croceipes (Cresson) (Hymenoptera: Braconidae), to detect volatile odors was successfully implemented in a previous study. To improve the odor-detecting ability of Wasp Hound®, searching behaviors of an individual wasp in a confined area are studied and modeled through stochastic differential equations in this paper. The wasps are conditioned to 20 mg of coffee when associated with food and subsequently, tested to 5, 10, 20, and 40 mg of coffee. A stochastic model is developed and validated based on three positive behavioral responses (walking, rotation around odor source, and self-rotation) from conditioned wasps at four different test dosages. The model is capable to reproducing the behaviors of conditioned wasps, and can be used to improve the ability of Wasp Hound® to assess changes in odor concentration. The model simulation results show the behaviors of conditioned wasps are significantly different when tested at different coffee dosages. We conjecture that the searching behaviors of conditioned wasps are based on the temporal and spatial neuron activity of olfactory receptor neurons and glomeruli, which are strongly correlated to the training dosages. The overall results demonstrate the utility of mathematical models for interpreting experimental observations, gaining novel insights into the dynamic behavior of classically conditioned wasps, as well as broadening the practical uses of Wasp Hound.

This study used a randomized control design to evaluate the effectiveness of AgTeen, an in-home, family-based farm safety intervention, in decreasing extra riding on tractors by youth. Having children as extra riders on tractors has deep roots in farm culture, but it can result in serious injury or death.The study randomized 151 families into 3 groups: parent-led intervention (fathers taught their families about farm safety), staff-led intervention (staff members who were peer farmers taught families), and a no-treatment control. Mothers, fathers, and all children aged 10-19 participated in the lessons.At study entry, 93% of youth reported that they had been an extra rider on a tractor in the past year. Although they were aware of the injury risk, fathers frequently gave tractor rides to their children. After the intervention, fathers in both AgTeen groups were less likely than control fathers to give youth tractor rides. Intervention youth were less likely than control youth to be extra riders. The intervention positively affected the extra-riding attitudes and injury risk perceptions of mothers and fathers. The parent-led and staff-led groups did not significantly differ across study outcomes.Findings confirm the effectiveness of a family-based intervention in decreasing extra riding on tractors by youth.