Keywords: Automation and Robotics in Agriculture, Agricultural Machinery Guidance and Control
Abstract: Field robotics applications typically require platforms to be deployed for extended periods of time, traversing large distances over often uneven terrains. To best leverage these platforms, it is important to know the energy costs of their operation, to maximise range and operating time. This is particularly the case for electrically powered robots, which share a single power source for actuation, sensing, and computation. In this paper we build upon our prior work---derivation of a physics-based power model for an omnidirectional wheeled mobile robot---to develop an energy-aware rough terrain motion planner, which enables efficient missions to be defined and carried out whilst considering the impact of the terrain slope and slippage characteristics on the energy cost of motion. We develop an effective pipeline for the generation of such missions on-the-fly, and compare the performance of the proposed energy-aware approach against a 3D Euclidean distance metric, experimentally verifying the increased energy efficiency of the resulting plans.

Keywords: Automation and Robotics in Agriculture, Agricultural Machinery Guidance and Control, Robust Control Systems for Agriculture
Abstract: Autonomous travel of agricultural machines/robots is achieved by movement along a straight line based on a map with the global positioning system (GPS). However, when farm produce is not grown in a straight line, the robots using GPS might damage the lettuce intended for harvesting. On solve this problem, autonomous travel using local information can improve the harvesting accuracy. Therefore, in this paper, we propose an autonomous travel method for a lettuce harvester with on-off actuators based on the relative positions of the lettuce and the harvest machine obtained via a camera. This study adopts model predictive control (MPC) to ensure that the vehicle follows the positions of some lettuce located ahead in the same line. To facilitate an application of an optimal method for determining an optimal input sequence in MPC, we transform a kinematics model of the vehicle into the time-state control form. The effectiveness of the proposed method is showed through numerical examples.

Keywords: Automation and Robotics in Agriculture, Agricultural Machinery Guidance and Control, Sensing and Automation in Animal Farming
Abstract: The task of bulldozing has many aspects that need more research before they can be fully automated. This paper presents a new method for calculating a path for vehicles tasked with pushing multiple objects to multiple goal locations. Unlike previous approaches which either focus on a single object, apply both push and pull operations, or operate in discrete domain, this solution works in a continuous space whilst handling multiple objects in a push-only scenario and doesn’t use computationally heavy methods to find a potential path planning solution. To achieve the results, two graph structures, one for the active vehicle and one for the passive pushable objects, is used. These graphs are then combined with an A* Search Algorithm to find a correct path for the bulldozer to take. The algorithm is tested on a Microban problem set which results in a 96% success rate.

Keywords: Automation and Robotics in Agriculture, Design and Control of Agricultural Implements, Agricultural Machinery Guidance and Control
Abstract: For an intelligent agricultural robot to reliably operate on a large-scale farm, it is crucial to accurately estimate its pose. In large outdoor environments, 3D LiDAR is a preferred sensor. Urban and agricultural scenarios are characteristically different, where the latter contains many poorly defined objects such as grass and trees with leaves that will generate noisy sensor signals. While state-of-the-art methods of state estimation using LiDAR, such as LiDAR odometry and mapping (LOAM), work well in urban scenarios, they will fail in the agricultural domain. Hence, we propose a mapping and localization system to cope with challenging agricultural scenarios. Our system maintains a high quality global map for subsequent reuses of relocalization or motion planning. This is beneficial as we avoid the unnecessary repetitively mapping process. Our experimental results show that we achieve comparable or better performance in state estimation, localization, and map quality when compared to LOAM.

Keywords: Automation and Robotics in Agriculture, Machine Vision and Robotics for Weed Control, Robust Control Systems for Agriculture
Abstract: The autonomous driving of agricultural machinery using information from global navigation satellite system (GNSS) information has developed rapidly because it is considered as a labor-saving measure in agriculture. The agricultural machinery is able to locate its position using a GNSS signal allowing it to move in an area autonomously. However, if machinery uses the GNSS signal only to self-locate it may run the risk of colliding with obstacles as it may not accurately sense the surrounding environment. Furthermore, sensors such as radars or lasers cannot distinguish between grass and obstacles; hence they cannot be used for sensing an agricultural environment including the detection of obstacles that are likely to be encountered by the machinery. Autonomous driving cannot be performed in environments such as orchards where the satellite positioning accuracy is low. This paper presents an autonomous driving system that we developed that is able to avoid obstacles and drive without the aid of a GNSS signal. The system uses an object detection system that is based on a stereo camera and deep learning technique i.e. convolutional neural networks as they can be used to recognize an environment and avoid obstacles. The autonomous driving ability of the vehicle was evaluated using real-time kinematic-GNSS to measure the true values through experiments that were conducted in the Tanashi Forest of the University of Tokyo.

Keywords: Internet of Things, Wireless Sensor Network, Sensing and Automation for Precision Irrigation
Abstract: With the development of technologies such as the Internet of Things and Big Data, agriculture is entering the digital age. Modern information technology and intelligent equipment are the development direction of agriculture. In this paper, a remote-control system for greenhouse environment was developed based on Raspberry Pi, MySQL and Android. The system consists of a greenhouse control system (subsystem A), an environment monitoring system (subsystem B), a MySQL remote database (subsystem C), and an Android monitoring software (subsystem D). The subsystem A is used to implement automatic fertigation and collect data from subsystem B, which will be uploaded to cloud server database (subsystem C). The subsystem B is used to monitor the environmental parameters in the greenhouse such as air temperature/humidity, soil temperature/humidity etc. The subsystem C is used for storage of greenhouse environment data and communication between subsystem B and subsystem D. The subsystem D is used for data visualization including real-time display, storage and analysis of greenhouse environmental data, and remote control of fertigation system including the EC/pH control of the fertigation solution, irrigation scheduling, and fertigation settings. A proportional-integral-derivative (PID) control algorithm is used for the fertigation process of the subsystem A. The low-power LoRa technology is used for subsystem B to achieve low-power long-distance data transmission, which is especially suitable to greenhouses. All monitoring settings can be operated remotely by users.

Keywords: Robust Control Systems for Agriculture, Sensing, Automation and Robotics in Plant Factory, Protected Cultivation and Greenhouses, Automation and Robotics in Agriculture
Abstract: This paper designs a controller of a passive nonlinear air-conditioning unit based on a loop-shaping approach, with respect of the environmental conditions for micro-climate in growth chambers and greenhouses. The control problem consists in mixing two prescribed air flows and tracking the mixed output set point temperature. Simulations were carried out on the closed-loop formed by connecting the controller to the nonlinear original system. The synthesized controller based on the H loop-shaping technique provides robust stability and performance. Using the -gap metric, the controller was shown to stabilize a large class of systems with uncertainties in power heating coefficients.

Keywords: Internet of Things, Decision Support Systems
Abstract: This work presents a new architecture to use greenhouse models as a service in an Internet of Thing (IoT) paradigm. Traditionally, models are implemented in specific programming environments for research purposes or coded within a software tool that can be used as Decision Support System (DSS) by companies or farmers. In the solution presented in this paper, the models are available in the cloud and they can be accessed as a service through Internet from any software tool or any device (computer, tablet or smartphone). On the other hand, the models can be accessed as DSS or as a simulator for control purposes. Examples for researcher and end-user points of view are presented for a tomato crop growth model.

Keywords: Pest and Disease Detection Management
Abstract: Greenhouse whitefly (Trialeurodes vaporariorum) is a major insect pest of greenhouse crops. To prevent the damage caused by whiteflies, farmers control the population of whiteflies by spraying pesticides in a regular basis. However, pesticides are costly and may affect the environment and health of farmers. To provide a more efficient way for applying pesticides, this research aims to develop a model for predicting the possible increase in whitefly population in greenhouses using autoregressive integrated moving average (ARIMA) and ARIMA with exogenous variables (ARIMAX). The data used in this work were collected using wireless imaging devices that can monitor the number of whiteflies trapped on sticky paper traps using an automatic insect counting algorithm. The wireless imaging devices were installed in a greenhouse that grew tomato seedlings, which is one of the host plants of whiteflies. The ARIMA and ARIMAX models were compared by setting different combinations of input data. Particularly, ARIMA includes only the whitefly count while ARIMAX includes the whitefly count and environmental data. Based on preliminary testing, the minimum number of input data was found to be at around 60 days to 90 days. ARIMAX was found to be the best model with input data including the increase in whitefly counts, temperature and humidity. In average, the RMSE for 7-day forecasting of the proposed method was found to be around 1.30. To assist farmers in decision-making for pesticide application scheduling, four levels of increase in whitefly count were defined such as Normal, Moderate, High, and Critical, which were determined using K-means clustering algorithm, and testing results on a testing dataset show an F1-scores of 0.86 and 0.42 for Normal and Moderate levels of daily increase in whitefly count.

Keywords: Design and Control of Agricultural Implements, Automation and Robotics in Agriculture
Abstract: Testing a seed drill indoors can save a great amount of time and labour force. Conventional seeding-performance test rigs usually need at least two people to operate it. Thus, a small size test rig with an automatic control system need to be developed. In this research, a control system was designed firstly. After that, rotation speed of seed-metering shaft were simulated by Simulink and the parameters of the control algorithm were obtained. Moreover, an experiment were conducted to calibrate the seed mass discharged in per cycle rotated by seed-metering shaft. According to the experimental results, the average seed mass in per cycle was 12.61 g. In each validating experiment, the value that the theoretical seed mass subtracted the actual one and then divided the theoretical seed mass was seen as the error percentage of seed mass. Furthermore, 100% subtracted the error percentage was viewed as the accuracy of seed mass. The results showed that all accuracies of seed mass were higher than 82%, and the average accuracy of total experiments was 89.12%. Furthermore, t-test results showed that there was not significant difference between the actual seed masses and the theoretical ones. According to the statistical results, we can conclude that the control system can be applied on seeding-performance test rigs with an desirable performance.

Keywords: Decision Support Systems, Automation and Robotics in Agriculture
Abstract: Autonomous systems operating in unstructured and complex agricultural environments are susceptible to errors leading to uncertain losses in the efficiency of the system. In robotic harvesting, these losses would translate into lower harvesting efficiency. To this end, it is desirable to improve harvesting efficiency of robotic systems through human collaboration. The added labor costs associated with human involvement could be a concern since the robotic harvesting systems are expected to reduce harvesting costs. Therefore, the objective of this work is to develop optimal human-robot collaboration policies that minimize the risk of economic losses by identifying the components of the system that need to be serviced by a human supervisor all while guaranteeing a desired level of financial return. The developed risk-averse optimization solution is verified in a simulated grove environment for Florida Valencia citrus.

Keywords: Big Data and Cloud Computing, Agricultural Machinery Guidance and Control, Decision Support Systems
Abstract: A scientific and reasonable recommendation algorithm for plant-protecting UAV (Unmanned Aerial Vehicle) operating helps UAV renters with optimal and economical solutions, as well improves the efficiency of plant protecting work and its management. This paper proposed a method of UAV operation recommendation for plant-protecting UAV users based on log data processing, and built a log data real-time analysis and calculation system using Apache Storm framework technology. According to the characteristics of the plant protection operations and the UAV users, two recommendation algorithms were designed and developed in this paper, one of them combined user feature and collaborative filtering and the other one was based on content filtering. In the former algorithm, the most similar neighbor users of the specific target user were investigated first, then the score of every similar neighbor’s each operation was calculated and an appropriate operation recommendation list was produced according to the calculation result of their weighted scores. The later recommendation algorithm took the target user’s interested operation location, spraying type and plant type into consideration, and it recommended the potential interested plant-protecting operations for the target user by analyzing his/her online search, browse history and records. These two algorithms were developed and integrated into an existing system we built earlier using Python. System test and analysis results showed that Storm framework could provide a real-time, low latency, high throughput and robust computing framework, and the log data processing based algorithms could give users the most meaningful plant-protecting operation recommendations according to the preset rules, which could potentially improve the system’s intellectuality, convenience and usability.

Keywords: Crop Monitoring, Soil, Plant and Environment Sensing, Crop Systems/Canopy Architectures, Breeding and Genetics for Precision and Automated Agriculture
Abstract: Root is one of the most important organs for plants to obtain water and nutrients so that its morphological research is critique for identifying plant growth conditions. Aiming at breakthrough of barriers and obtaining accurate root phenotype data based on the original plant root image, a method of Arabidopsis thaliana root image restoration based on GAN (generative adversarial network) was proposed in this paper. Firstly, a second generation Kinect camera is used to capture the matched data set for training the GAN, which includes high-resolution images of some objects and their matched fuzzy and distort images, and high-resolution images of Arabidopsis’ roots and their images in the biogel. Secondly, a GAN with attention mechanism is constructed and trained. The network mainly consists of two parts: the generator and the discriminator with attention mechanism. It is multi-layer convolution network, except that the generator adopts a de-convolution structure to carry out the super-resolution reconstruction. The generator is responsible for converting a fuzzy image into high-resolution image, and the discriminator is used to distinguish whether the inputted image is derived from the prepared dataset or generated by the generator. With the progress of network training, the generator is getting better and better at generating images, the same is true for the effect of the discriminator discriminating the image, that is, the better mapping relationship between the blurred or partially missing image and the high resolution complete image is established. Finally, import the root image of the Arabidopsis planted in the biogel into the trained network and the repaired and restored root image can be obtained. Compared with the original image, the restored one has more accurate details and accordingly more accurate root morphology parameters are computed. The experiment results showed that the proposed method can be used to achieve the super-resolution reconstruction and complete the incomplete or blur Arabidopsis root images.

Keywords: Crop Monitoring, Soil, Plant and Environment Sensing, Precision Agriculture and Variable Rate Technologies
Abstract: The function of roots is exceedingly important for crops and the observation of root growth is essential for understanding the plant physiology and for the improvement of crops productivity. The current study focused on the use of sound waves which can propagate in water and soil and are expected to interact with roots. In order to monitor the growth of roots non-invasively, a prototype sound-based root growth detection system was developed, and its performance was evaluated on Komatsuna. The attenuation amount of transmitted sound intensity increased during the cultivation period, and a positive correlation between the attenuation amount and roots growth was observed. Based on this observation, a sound-based root growth detection system has the possibility to indirectly detect root growth by an extremely low-cost and simple method.

Keywords: Sensing and Automation with UAVs, Precision Agriculture and Variable Rate Technologies, Automation and Robotics in Agriculture
Abstract: Precision agriculture is a mode of modern agricultural production and management on the basis of information technology, and it is an important way to achieve low consumption, high efficiency, fine quality and safety in agriculture. Moreover, accurate recognition and extraction of field operation area (FOA) boundary is a crucial basic data to implement precision agriculture. Due to the irregular shape, different planting ways and inconsistent size of FOA, it is difficult to recognize and extract the boundary. Therefore, the boundary of FOA based on UAV remote sensing images was classified in this paper, and a priori rule was proposed according to the boundary characteristics. Combined with the classical algorithm of image processing, a boundary recognition and extraction system was developed by LabVIEW to obtain the effective boundary of FOA. At last, the boundary recognition method and the accuracy and real-time performance of the extraction system were tested, and the result shows the method and system can accurately recognize and extract the actual boundary of FOA, and they are adapted to three different types of boundaries and image resolutions. The system has good real-time performance, and the single-frame image processing time does not exceed 100ms when the image resolution is lower than 1920*1280.

Keywords: Automation and Robotics in Agriculture, Sensing and Automation with UAVs, Agricultural Machinery Guidance and Control
Abstract: Quadcopters can play an important role in precision agriculture; mostly having been researched and commercialized for the open field conditions. While greenhouse applications have been limited by the positioning system. Conventional GPS positioning systems are difficult to use in greenhouses. Spread Spectrum Sound-based Positioning System (SSSPS) on the other hand, has the potential to be used on a quadcopter in a greenhouse. In this research, we evaluated whether acoustic noise from a quadcopter over wide frequency range would interfere with the SSSPS operation. We set a microphone on the quadcopter, and evaluated the SSSPS system for quadcopter noise tolerance with several signals. The results demonstrate the navigation system can tolerate quadcopter noise over distances up to 15 m between the tweeter and the microphone on the quadcopter with the RMSE of 5.9 mm.

Keywords: Sensing and Automation with UAVs, Automation and Robotics in Agriculture, Soil, Plant and Environment Sensing
Abstract: Globally, large-scale irrigation canal networks serve as the backbone of agriculture in many important river basins. However, these water channels are in a constant threat of erosion, silt accumulation and structural damages over time which significantly reduces the water carrying capacity. Therefore, periodic inspections of the canals are required for critical operations and maintenance tasks. Due to the vast lengths of the channels and time-critical operations, automation has become a necessity. In this paper, we have proposed an aerial autonomous canal traversal system using ResNet50 inspired deep convolutional neural network. Given the uniqueness of our problem, we have generated our dataset for supervised learning and validation and later evaluated the proposed approach on a real canal. We have implemented our approach on a COTS micro-aerial vehicle. We have designed our system in such a way that it takes 200ms from perception to action thereby making the system real-time. We compare the superior performance of our ResNet50 inspired network with other state-of-the-art CNNs trained on canal datasets.

Keywords: Sensing and Automation with UAVs, Automation and Robotics in Agriculture, Robust Control Systems for Agriculture
Abstract: A distributed control strategy of multi quadrotor UAV systems is proposed in this work to address area coverage in precision agriculture. It adopts a region-based control approach by restricting the motion of UAVs within a desired dynamical region with planned dynamics while providing collision avoidance between UAVs. Voronoi partitions that can be computed in a distributed way are used to determine UAVs positions within the moving region which provides a robust and scalable solution. Simulations are done using Gazebo and Robot Operating System (ROS) to evaluate the performance of the proposed method showing good results.

Keywords: Pest and Disease Detection Management, Sensing, Automation and Robotics for Post-Harvest/Processing
Abstract: This work focuses on the evaluation of growth characteristics of two kinds of Colletotrichum species, i.e., Colletotrichum truncatum and Colletotrichum gloeosporioides, by near infrared (NIR) hyperspectral imaging. The hyperspectral images of the two fungi growing on potato ager medium were recorded daily for 6 days. The average spectra of each fungi were extracted, and the reflectance of the average spectra preliminarily indicated the growth phases of fungi. Principal component analysis (PCA) and support vector machine classifier (SVM) were applied on the full spectral range. Two groups of optimal PCs (PC1-5 for C. truncatum; and PC1-2, PC4 and PC6 for C. gloeosporioides) were respectively selected by Wilks-λ criterion to build the PCA-SVM classification models. The identification accuracies were 90.83% and 94.17% for C. truncatum and C. gloeosporioides, respectively. To simplify the prediction models, competitive adaptive reweighted sampling (CARS) was employed to choose optimal wavelengths. Total twelve (471.8, 597.4, 777.6, 790.2, 792.8, 795.3, 861.8, 882.5, 892.9, 895.5, 898.1, 963.6 nm) and ten (516.3, 523.4, 563.8, 571.0, 747.4, 802.9, 825.9, 828.4, 831.0, 856.7 nm) wavelengths were selected for the two Colletotrichum species. Corresponding SVM models build by those wavelengths could identify fungal growth days with the accuracies of 97.50%. Results indicate that NIR hyperspectral imaging is a powerful tool to evaluate the growth characteristics of C. truncatum and C. gloeosporioides.

Keywords: Soil, Plant and Environment Sensing, Crop Monitoring
Abstract: To find a convenient and non-destructive way to detect the water stress status of maize plant, hyperspectral technology was used to detect maize leaf water content (LWC) in this paper. Experiments were conducted on maize at the seedling stage in Beijing, China. Firstly, hyperspectral images of 85 maize plant leaves were obtained. To remove redundant information, two methods were used for dimensionality reduction, which are Principal component analysis (PCA) and kullback-leibler divergence (KLD) methods, resulting in 10 and 6 wavebands, respectively. These wavebands were then used to establish water stress detection model based on support vector regress (SVR) method. Particle swarm optimization (PSO) algorithm was applied to optimize the parameters for the support vector regression (SVR), including the penalty factor C and kernel function parameter g. Finally, the SVR model with the optimized parameters were trained with 65 pretreatment samples, and the generalization ability of the model was tested with the remaining 20 samples. Experimental results show the model built by bands selected by KLD method (R=0.7684) is better than PCA method(R=0.3513).Future research should aim at testing the model generation for different growing stage of maize.

Keywords: Soil, Plant and Environment Sensing, Precision Agriculture and Variable Rate Technologies, Automation and Robotics in Agriculture
Abstract: Soil total nitrogen is an important indicator of soil fertility. In order to achieve a general applicability of the soil total nitrogen content detector developed based on the principle of spectroscopy, this paper selects the universal nitrogen-sensitive wavelength of soil suitable for different types of soil and models optimization. Firstly, the spectral characteristic curves of black soil, cinnamon soil and tidal soil were measured, the analysis found that under the same soil total nitrogen content, the absorbance curves of different types of soils are quite different. Then, the Monte Carlo non-information variable elimination (MC-UVE) algorithm was used to screen the soil total nitrogen sensitive wavelengths of the three soil types, the common sensitive spectral regions of different soil types were screened, and the spectral wavelength ranges were 895-911 nm, 1047-1065 nm, 1211-1232 nm, 1468-1482 nm, 1691-1699 nm and 2095-2109 nm. On the basis of the selected spectrum, the continuous projection algorithm was used to optimize the sensitive wavelength again, and the characteristic wavelengths suitable for the prediction of three soil types were 902, 1054, 1221, 1478, 1697, 1969, 2104 nm. The prediction model was established by the absorbance values of the soil total sensitive wavelength and the soil total nitrogen standard value, the highest model accuracy was obtained from the Support Vector Regression model of different soil type. The calibration RC2 of the black soil model was 0.922 and the verification RV2 of the black model was 0.892. The calibration RC2 of the cinnamon soil model was 0.905 and the verification RV2 of the cinnamon model was 0.877. The calibration RC2 of the tidal soil model was 0.892 and the verification RV2 of the tidal model was 0.873. The experimental results show that the common sensitive wavelengths of different soils can be used as a reference for the selection of light source and filter for soil total nitrogen content detector. However, the soil total nitrogen model needs to consider parameter setting for different soil types when modelling different soils to achieve universal applicability of the detector for different soil types.

Keywords: Soil, Plant and Environment Sensing, Sensing, Automation and Robotics for Post-Harvest/Processing, Machine Vision and Robotics for Crop Harvesting
Abstract: Imaging beyond RGB bands has the ability to provide solutions for non-destructive detection of objects or scene properties. Methods often used in research include hyperspectral cameras and spectrometers; the former is expensive and the latter can only provide point measurements. Multispectral imaging can provide low-cost solutions in instances where RGB does not suffice, however, evaluation of filters used in multispectral sensors or the use of off-the-shelf filters for user specific applications have not been explicitly analysed. This paper proposes a novel method for using spectrometer data by firstly limiting the search space to existing off-the-shelf filters, modelling those filters, and then applying the desired model for classification or regression to more appropriately model multispectral imaging performance. The results indicate that point measurements produced by spectrometers may not correspond with the response achieved when using off-the-shelf filters; particularly in the instance where the spectral profile of the measured object varies significantly within the transmission regions of the filter. The method presented focuses on the practical estimation of filter performance when selecting filters.