Abstract Modern agriculture relies heavily on the precise application of chemicals such as fertilisers, herbicides, and pesticides, which directly affect both crop yield and environmental footprint. Therefore, it is crucial to assess the accuracy of precision sprayers regarding the spatial location of spray deposits. However, there is currently no fully automated evaluation method for this. In this study, we collected a novel dataset from a precision spot spraying system to enable us to classify and detect spray deposits on target weeds and non-target crops. We employed multiple deep convolutional backbones for this task; subsequently, we have proposed a robustness testing methodology for evaluation purposes. We experimented with two novel data augmentation techniques: subtraction and thresholding which enhanced the classification accuracy and robustness of the developed models. On average, across nine different tests and four distinct convolutional neural networks, subtraction improves robustness by 50.83%, and thresholding increases by 42.26% from a baseline. Additionally, we have presented the results from a novel weakly supervised object detection task using our dataset, establishing a baseline Intersection over Union score of 42.78%. Our proposed pipeline includes an explainable artificial intelligence stage and provides insights not only into the spatial location of the spray deposits but also into the specific filtering methods within that spatial location utilised for classification.

Plants are dynamic organisms and understanding temporal variations in vegetation is an essential problem for robots in the wild. However, associating repeated 3D scans of plants across time is challenging. A key step in this process is re-identifying and tracking the same individual plant components over time. Previously, this has been achieved by comparing their global spatial or topological location. In this work, we demonstrate how using shape features improves temporal organ matching. We present a landmark-free shape compression algorithm, which allows for the extraction of 3D shape features of leaves, characterises leaf shape and curvature efficiently in few parameters, and makes the association of individual leaves in feature space possible. The approach combines 3D contour extraction and further compression using Principal Component Analysis (PCA) to produce a shape space encoding, which is entirely learned from data and retains information about edge contours and 3D curvature. Our evaluation on temporal scan sequences of tomato plants shows, that incorporating shape features improves temporal leaf-matching. A combination of shape, location, and rotation information proves most informative for recognition of leaves over time and yields a true positive rate of 75%, a 15% improvement on sate-of-the-art methods. This is essential for robotic crop monitoring, which enables whole-of-lifecycle phenotyping.

Tractors and heavy machinery have been used for decades to improve the quality and overall agriculture production. Moreover, agriculture is becoming a trend domain for robotics, and as a consequence, the efforts towards automatizing agricultural task increases year by year. However, for autonomous applications, accident prevention is of prior importance for warrantying human safety during operation in any scenario. This paper rephrases human safety as a classification problem using a custom distance criterion where each detected human gets a risk level classification. We propose the use of a neural network trained to detect and classify humans in the scene according to these criteria. The proposed approach learns from real-world data corresponding to an open-field scenario and is assessed with a custom risk assessment method.

Data augmentation can be a simple yet powerful tool for autonomous robots to fully utilise available data for self-supervised identification of atypical scenes or objects. State-of-the-art augmentation methods arbitrarily embed “structural” peculiarity on typical images so that classifying these artefacts can provide guidance for learning representations for the detection of anomalous visual signals. In this paper, however, we argue that learning such structure-sensitive representations can be a suboptimal approach to some classes of anomaly (e.g., unhealthy fruits) which could be better recognised by a different type of visual element such as “colour”. We thus propose Channel Randomisation as a novel data augmentation method for restricting neural networks to learn encoding of “colour irregularity” whilst predicting channel-randomised images to ultimately build reliable fruit-monitoring robots identifying atypical fruit qualities. Our experiments show that (1) this colour-based alternative can better learn representations for consistently accurate identification of fruit anomalies in various fruit species, and also, (2) unlike other methods, the validation accuracy can be utilised as a criterion for early stopping of training in practice due to positive correlation between the performance in the self-supervised colour-differentiation task and the subsequent detection rate of actual anomalous fruits. Also, the proposed approach is evaluated on a new agricultural dataset, Riseholme-2021, consisting of 3.5K strawberry images gathered by a mobile robot, which we share online to encourage active agri-robotics research.

This paper introduces a new method for robot motion planning and navigation in uneven environments through a surfel representation of underlying point clouds. The proposed method addresses the shortcomings of state-of-the-art navigation methods by incorporating both kinematic and physical constraints of a robot with standard motion planning algorithms (e.g., those from the Open Motion Planning Library), thus enabling efficient sampling-based planners for challenging uneven terrain navigation on raw point cloud maps. Unlike techniques based on Digital Elevation Maps (DEMs), our novel surfel-based state-space formulation and implementation are based on raw point cloud maps, allowing for the modeling of overlapping surfaces such as bridges, piers, and tunnels. Experimental results demonstrate the robustness of the proposed method for robot navigation in real and simulated unstructured environments. The proposed approach also optimizes planners' performances by boosting their success rates up to 5x for challenging unstructured terrain planning and navigation, thanks to our surfel-based approach's robot constraint-aware sampling strategy. Finally, we provide an open-source implementation of the proposed method to benefit the robotics community.

Understanding how people are likely to move is key to efficient and safe robot navigation in human environments. However, mobile robots can only observe a fraction of the environment at a time, while the activity patterns of people may also change at different times. This article introduces a new methodology for mobile robot exploration to maximize the knowledge of human activity patterns by deciding where and when to collect observations. We introduce an exploration policy driven by the entropy levels in a spatio-temporal map of pedestrian flows, and compare multiple spatio-temporal exploration strategies including both informed and uninformed approaches. The evaluation is performed by simulating mobile robot exploration using real sensory data from three long-term pedestrian datasets. The results show that for certain scenarios the models built with proposed exploration system can better predict the flow patterns than uninformed strategies, allowing the robot to move in a more socially compliant way, and that the exploration ratio is a key factor when it comes to the model prediction accuracy.

Safety integration components for robotic applications are a mandatory feature for any autonomous mobile application, including human avoidance behaviors. This paper proposes a novel parametrizable scene risk evaluator for open-field applications that use humans motion predictions and pre-defined hazard zones to estimate a braking factor. Parameters optimization uses simulated data. The evaluation is carried out by simulated and real-time scenarios, showing the impact of human predictions in favor of risk reductions on agricultural applications.

Recent developments in agriculture have high-lighted the potential of as well as the need for the use of robotics. Various processes in this field can benefit from the proper use of state of the art technology [1], in terms of efficiency as well as quality. One of these areas is the harvesting of ripe fruit.In order to be able to automate this process, a robotic harvester needs to be aware of the full poses of the crop/fruit to be collected in order to perform proper path- and collision-planning. The current state of the art mainly considers problems of detection and segmentation of fruit with localisation limited to the 3D position only. The reliable and real-time estimation of the respective orientations remains a mostly unaddressed problem.In this paper, we present a compact and efficient network architecture for estimating the orientation of soft fruit such as strawberries from colour and, optionally, depth images. The proposed system can be automatically trained in a realistic simulation environment. We evaluate the system’s performance on simulated datasets and validate its operation on publicly available images of strawberries to demonstrate its practical use. Depending on the amount of training data used, coverage of state space, as well as the availability of RGB-D or RGB data only, mean errors of as low as 11° could be achieved.

Global navigation satellite system (GNSS) has been considered as a panacea for positioning and tracking since the last decade. However, it suffers from severe limitations in terms of accuracy, particularly in highly cluttered and indoor environments. Though real-time kinematics (RTK) supported GNSS promises extremely accurate localisation, employing such services are expensive, fail in occluded environments and are unavailable in areas where cellular base stations are not accessible. It is, therefore, necessary that the GNSS data is to be filtered if high accuracy is required. Thus, this article presents a GNSS-based particle filter that exploits the spatial constraints imposed by the environment. In the proposed setup, the state prediction of the sample set follows a restricted motion according to the topological map of the environment. This results in the transition of the samples getting confined between specific discrete points, called the topological nodes, defined by a topological map. This is followed by a refinement stage where the full set of predicted samples goes through weighting and resampling, where the weight is proportional to the predicted particle's proximity with the GNSS measurement. Thus, a discrete space continuous-time Bayesian filter is proposed, called the Topological Particle Filter (TPF).The proposed TPF is put to test by localising and tracking fruit pickers inside polytunnels. Fruit pickers inside polytunnels can only follow specific paths according to the topology of the tunnel. These paths are defined in the topological map of the polytunnels and are fed to TPF to tracks fruit pickers. Extensive datasets are collected to demonstrate the improved discrete tracking of strawberry pickers inside polytunnels thanks to the exploitation of the environmental constraints.

Real-time 3D perception of the environment is crucial for the adoption and deployment of reliable autonomous harvesting robots in agriculture. Using data collected with RGB-D cameras under farm field conditions, we present two methods for processing 3D data that reliably detect mature broccoli heads. The proposed systems are efficient and enable real-time detection on depth data of broccoli crops using the organised structure of the point clouds delivered by a depth sensor. The systems are tested with datasets of two broccoli varieties collected in planted fields from two different countries. Our evaluation shows the new methods outperform state-of-the-art approaches for broccoli detection based on both 2D vision-based segmentation techniques and depth clustering using the Euclidean proximity of neighbouring points. The results show the systems are capable of accurately detecting the 3D locations of broccoli heads relative to the vehicle at high frame rates.

Models of human behaviour, such as pedestrian flows, are beneficial for safe and efficient operation of mobile robots. We present a new methodology for benchmarking of pedestrian flow models based on the afforded safety of robot navigation in human-populated environments. While previous evaluations of pedestrian flow models focused on their predictive capabilities, we assess their ability to support safe path planning and scheduling. Using real-world datasets gathered continuously over several weeks, we benchmark state-of-the-art pedestrian flow models, including both time-averaged and time-sensitive models. In the evaluation, we use the learned models to plan robot trajectories and then observe the number of times when the robot gets too close to humans, using a predefined social distance threshold. The experiments show that while traditional evaluation criteria based on model fidelity differ only marginally, the introduced criteria vary significantly depending on the model used, providing a natural interpretation of the expected safety of the system. For the time-averaged flow models, the number of encounters increases linearly with the percentage operating time of the robot, as might be reasonably expected. By contrast, for the time-sensitive models, the number of encounters grows sublinearly with the percentage operating time, by planning to avoid congested areas and times.

Understanding how people are likely to behave in an environment is a key requirement for efficient and safe robot navigation. However, mobile platforms are subject to spatial and temporal constraints, meaning that only partial observations of human activities are typically available to a robot, while the activity patterns of people in a given environment may also change at different times. To address these issues we present as the main contribution an exploration strategy for acquiring models of pedestrian flows, which decides not only the locations to explore but also the times when to explore them. The approach is driven by the uncertainty from multiple Poisson processes built from past observations. The approach is evaluated using two long-term pedestrian datasets, comparing its performance against uninformed exploration strategies. The results show that when using the uncertainty in the exploration policy, model accuracy increases, enabling faster learning of human motion patterns.

Precise knowledge of pose is of great importance for reliable operation of mobile robots in outdoor environments. Simultaneous localization and mapping (SLAM) is the online construction of a map during exploration of an environment. One of the components of SLAM is loop closure detection, identifying that the same location has been visited and is present on the existing map, and localizing against it. We have shown in previous work that using semantics from a deep segmentation network in conjunction with the Normal Distributions Transform (NDT) point cloud registration improves the robustness, speed and accuracy of lidar odometry. In this work we extend the method for loop closure detection, using the labels already available from local registration into NDT Histograms, and we present a SLAM pipeline based on Semantic assisted NDT and PointNet ++. We experimentally demonstrate on sequences from the KITTI benchmark that the map descriptor we propose outperforms NDT Histograms without semantics, and we validate its use on a SLAM task.

Point cloud registration is a core problem of many robotic applications, including simultaneous localization and mapping. The Normal Distributions Transform (NDT) is a method that fits a number of Gaussian distributions to the data points, and then uses this transform as an approximation of the real data, registering a relatively small number of distributions as opposed to the full point cloud. This approach contributes to NDT's registration robustness and speed but leaves room for improvement in environments of limited structure. To address this limitation we propose a method for the introduction of semantic information extracted from the point clouds into the registration process. The paper presents a large scale experimental evaluation of the algorithm against NDT on two publicly available benchmark data sets. For the purpose of this test a measure of smoothness is used for the semantic partitioning of the point clouds. The results indicate that the proposed method improves the accuracy, robustness and speed of NDT registration, especially in unstructured environments, making NDT suitable for a wider range of applications.

In this paper we present our work on the indoor positioning of users (shoppers), using a network of Bluetooth Low Energy (BLE) beacons deployed in a large wholesale shopping store. Our objective is to accurately determine which product sections a user is adjacent to while traversing the store, using RSSI readings from multiple beacons, measured asynchronously on a standard commercial mobile device. We further wish to leverage the store layout (which imposes natural constraints on the movement of users) and the physical configuration of the beacon network, to produce a robust and efficient solution. We introduce our node-graph model of user location, which is designed to represent the location layout. We also present our experimental work which includes an investigation of signal characteristics along and across aisles. We propose three methods of localization, using a “nearest-beacon” approach as a base-line; exponentially averaged weighted range estimates; and a particle-filter method based on the RSSI attenuation model and Gaussian-noise. Our results demonstrate that the particle filter method significantly out-performs the others. Scalability also makes this method ideal for applications run on mobile devices with more limited computational capabilities.

This paper presents a 3D vision system for robotic harvesting of broccoli using low-cost RGB-D sensors. The presented method addresses the tasks of detecting mature broccoli heads in the field and providing their 3D locations relative to the vehicle. The paper evaluates different 3D features, machine learning and temporal filtering methods for detection of broccoli heads. Our experiments show that a combination of Viewpoint Feature Histograms, Support Vector Machine classifier and a temporal filter to track the detected heads results in a system that detects broccoli heads with 95.2% precision. We also show that the temporal filtering can be used to generate a 3D map of the broccoli head positions in the field.

This paper presents a new approach to mobile robot mapping in long-term scenarios. So far, the environment models used in mobile robotics have been tailored to capture static scenes and dealt with the environment changes by means of `memory decay'. While these models keep up with slowly changing environments, their utilization in dynamic, real world environments is difficult. The representation proposed in this paper models the environment's spatio-temporal dynamics by its frequency spectrum. The spectral representation of the time domain allows to identify, analyse and remember regularly occurring environment processes in a computationally efficient way. Knowledge of the periodicity of the different environment processes constitutes the model predictive capabilities, which are especially useful for long-term mobile robotics scenarios. In the experiments presented, the proposed approach is applied to data collected by a mobile robot patrolling an indoor environment over a period of one week. Three scenarios are investigated, including intruder detection and 4D mapping. The results indicate that the proposed method allows to represent arbitrary timescales with constant (and low) memory requirements, achieving compression rates up to 106. Moreover, the representation allows for prediction of future environment states with ~ 90% precision.

This paper presents an approach for tracking multiple persons using a combination of colour and thermal vision sensors on a mobile robot. First, an adaptive colour model is incorporated into the measurement model of the tracker. Second, a new approach for detecting occlusions is introduced, using a machine learning classifier for pairwise comparison of persons (classifying which one is in front of the other). Third, explicit occlusion handling is then incorporated into the tracker.

In this paper we present a vision-based approach to detect, track and identify people on a mobile robot in real time. While most vision systems for tracking people on mobile robots use skin color information, we present an approach using thermal images and a fast contour model together with a particle filter. With this method a person can be detected independently from current light conditions and in situations where no skin color is visible (the person is not close or does not face the robot). Tracking in thermal images is used as an attention system to get an estimate of the position of a person. Based on this estimate we use a pan-tilt camera to zoom to the expected face region and apply a fast face tracker in combination with face recognition to identify the person.