Salah Sukkarieh

He Kong, Salah Sukkarieh

This paper considers a practical scenario where a classical estimation method might have already been implemented on a certain platform when one tries to apply more advanced techniques such as moving horizon estimation (MHE). We are interested to utilize MHE to upgrade, rather than completely discard, the existing estimation technique. This immediately raises the question how one can improve the estimation performance gradually based on the pre-estimator. To this end, we propose a general methodology which incorporates the pre-estimator with a tuning parameter λ between 0 and 1 into the quadratic cost functions that are usually adopted in MHE. We examine the above idea in two standard MHE frameworks that have been proposed in the existing literature. For both frameworks, when λ = 0, the proposed strategy exactly matches the existing classical estimator; when the value of λ is increased, the proposed strategy exhibits a more aggressive normalized forgetting effect towards the old data, thereby increasing the estimation performance gradually.

Jasper Brown, Cameron Clark, Sabrina Lomax et al.

Modern livestock farming is increasingly data driven and frequently relies on efficient remote sensing to gather data over wide areas. High resolution satellite imagery is one such data source, which is becoming more accessible for farmers as coverage increases and cost falls. Such images can be used to detect and track animals, monitor pasture changes, and understand land use. Many of the data driven models being applied to these tasks require ground truthing at resolutions higher than satellites can provide. Simultaneously, there is a lack of available aerial imagery focused on farmland changes that occur over days or weeks, such as herd movement. With this goal in mind, we present a new multi-temporal dataset of high resolution UAV imagery which is artificially degraded to match satellite data quality. An empirical blurring metric is used to calibrate the degradation process against actual satellite imagery of the area. UAV surveys were flown repeatedly over several weeks, for specific farm locations. This 5cm/pixel data is sufficiently high resolution to accurately ground truth cattle locations, and other factors such as grass cover. From 33 wide area UAV surveys, 1869 patches were extracted and artificially degraded using an accurate satellite optical model to simulate satellite data. Geographic patches from multiple time periods are aligned and presented as sets, providing a multi-temporal dataset that can be used for detecting changes on farms. The geo-referenced images and 27,853 manually annotated cattle labels are made publicly available.

Nicholas Harrison, Nathan Wallace, Salah Sukkarieh

The efficient collection of samples is an important factor in outdoor information gathering applications on account of high sampling costs such as time, energy, and potential destruction to the environment. Utilization of available a-priori data can be a powerful tool for increasing efficiency. However, the relationships of this data with the quantity of interest are often not known ahead of time, limiting the ability to leverage this knowledge for improved planning efficiency. To this end, this work combines transfer learning and active learning through a Multi-Task Gaussian Process and an information-based objective function. Through this combination it can explore the space of hypothetical inter-quantity relationships and evaluate these hypotheses in real-time, allowing this new knowledge to be immediately exploited for future plans. The performance of the proposed method is evaluated against synthetic data and is shown to evaluate multiple hypotheses correctly. Its effectiveness is also demonstrated on real datasets. The technique is able to identify and leverage hypotheses which show a medium or strong correlation to reduce prediction error by a factor of 1.4--3.4 within the first 7 samples, and poor hypotheses are quickly identified and rejected eventually having no adverse effect.

Jasper Brown, Yongliang Qiao, Cameron Clark et al.

Knowing where livestock are located enables optimized management and mustering. However, Australian farms are large meaning that many of Australia's livestock are unmonitored which impacts farm profit, animal welfare and the environment. Effective animal localisation and counting by analysing satellite imagery overcomes this management hurdle however, high resolution satellite imagery is expensive. Thus, to minimise cost the lowest spatial resolution data that enables accurate livestock detection should be selected. In our work, we determine the association between object detector performance and spatial degradation for cattle, sheep and dogs. Accurate ground truth was established using high resolution drone images which were then downsampled to various ground sample distances (GSDs). Both circular and cassegrain aperture optics were simulated to generate point spread functions (PSFs) corresponding to various optical qualities. By simulating the PSF, rather than approximating it as a Gaussian, the images were accurately degraded to match the spatial resolution and blurring structure of satellite imagery. Two existing datasets were combined and used to train and test a YoloV5 object detection network. Detector performance was found to drop steeply around a GSD of 0.5m/px and was associated with PSF matrix structure within this GSD region. Detector mAP performance fell by 52 percent when a cassegrain, rather than circular, aperture was used at a 0.5m/px GSD. Overall blurring magnitude also had a small impact when matched to GSD, as did the internal network resolution. Our results here inform the selection of remote sensing data requirements for animal detection tasks, allowing farmers and ecologists to use more accessible medium resolution imagery with confidence.

Jennifer Wakulicz, He Kong, Salah Sukkarieh

Trajectory optimization of sensing robots to actively gather information of targets has received much attention in the past. It is well-known that under the assumption of linear Gaussian target dynamics and sensor models the stochastic Active Information Acquisition problem is equivalent to a deterministic optimal control problem. However, the above-mentioned assumptions regarding the target dynamic model are limiting. In real-world scenarios, the target may be subject to disturbances whose models or statistical properties are hard or impossible to obtain. Typical scenarios include abrupt maneuvers, jumping disturbances due to interactions with the environment, anomalous misbehaviors due to system faults/attacks, etc. Motivated by the above considerations, in this paper we consider targets whose dynamic models are subject to arbitrary unknown inputs whose models or statistical properties are not assumed to be available. In particular, with the aid of an unknown input decoupled filter, we formulate the sensor trajectory planning problem to track evolution of the target state and analyse the resulting performance for both the state and unknown input evolution tracking. Inspired by concepts of Reduced Value Iteration, a suboptimal solution that expands a search tree via Forward Value Iteration with informativeness-based pruning is proposed. Concrete suboptimality performance guarantees for tracking both the state and the unknown input are established. Numerical simulations of a target tracking example are presented to compare the proposed solution with a greedy policy.

Stuart Eiffert, Nathan D. Wallace, He Kong et al.

For mobile robots to be effectively applied to real world unstructured environments -- such as large scale farming -- they require the ability to generate adaptive plans that account both for limited onboard resources, and the presence of dynamic changes, including nearby moving individuals. This work provides a real world empirical evaluation of our proposed hierarchical framework for long-term autonomy of field robots, conducted on University of Sydney's Swagbot agricultural robot platform. We demonstrate the ability of the framework to navigate an unstructured and dynamic environment in an effective manner, validating its use for long-term deployment in large scale farming, for tasks such as autonomous weeding in the presence of moving individuals.

Stuart Eiffert, Nathan D. Wallace, He Kong et al.

Achieving long-term autonomy for mobile robots operating in real-world unstructured environments such as farms remains a significant challenge. This is made increasingly complex in the presence of moving humans or livestock. These environments require a robot to be adaptive in its immediate plans, accounting for the state of nearby individuals and the response that they might have to the robot's actions. Additionally, in order to achieve longer-term goals, consideration of the limited on-board resources available to the robot is required, especially for extended missions such as weeding an agricultural field. To achieve efficient long-term autonomy, it is thus crucial to understand the impact that online dynamic updates to an energy efficient offline plan might have on resource usage whilst navigating through crowds or herds. To address these challenges, a hierarchical planning framework is proposed, integrating an online local dynamic path planner with an offline longer-term objective-based planner. This framework acts to achieve long-term autonomy through awareness of both dynamic responses of individuals to a robot's motion and the limited resources available. This paper details the hierarchical approach and its integration on a robotic platform, including a comprehensive description of the planning framework and associated perception modules. The approach is evaluated in real-world trials on farms, requiring both consideration of limited battery capacity and the presence of nearby moving individuals. These trials additionally demonstrate the ability of the framework to adapt resource use through variation of the local dynamic planner, allowing adaptive behaviour in changing environments. A summary video is available at https://youtu.be/DGVTrYwJ304.

Jasper Brown, Salah Sukkarieh

Many automated operations in agriculture, such as weeding and plant counting, require robust and accurate object detectors. Robotic fruit harvesting is one of these, and is an important technology to address the increasing labour shortages and uncertainty suffered by tree crop growers. An eye-in-hand sensing setup is commonly used in harvesting systems and provides benefits to sensing accuracy and flexibility. However, as the hand and camera move from viewing the entire trellis to picking a specific fruit, large changes in lighting, colour, obscuration and exposure occur. Object detection algorithms used in harvesting should be robust to these challenges, but few datasets for assessing this currently exist. In this work, two new datasets are gathered during day and night operation of an actual robotic plum harvesting system. A range of current generation deep learning object detectors are benchmarked against these. Additionally, two methods for fusing depth and image information are tested for their impact on detector performance. Significant differences between day and night accuracy of different detectors is found, transfer learning is identified as essential in all cases, and depth information fusion is assessed as only marginally effective. The dataset and benchmark models are made available online.

Jasper Brown, Salah Sukkarieh

The human labour required for tree crop harvesting is a major cost component in fruit production and is increasing. To address this, many existing research works have sought to demonstrate commercially viable robotic harvesting for tree crops, though successful commercial products resulting from these have been few and far between. Systems developed for specific crops such as sweet peppers or apples have shown promise, but the vast majority of cultivar types remain unaddressed and developing a specific system for each one is inefficient. In this work a flexible and modular development platform is presented, this can be used to test specific design choices on different fruit and growing conditions. The system is evaluated in a commercial plum orchard, with no crop modifications. Some existing techniques are found to be counterproductive for plums, while soft robotics and persistent target tracking significantly improve performance. A harvest success rate of 42% was observed, with lower than expected effectiveness based on prior testing with apples. The plum type and growing style pose unique challenges which are examined in the context of system module design choices.

Stuart Eiffert, Nathan D. Wallace, He Kong et al.

Achieving long term autonomy of robots operating in dynamic environments such as farms remains a significant challenge. Arguably, the most demanding factors to achieve this are the on-board resource constraints such as energy, planning in the presence of moving individuals such as livestock and people, and handling unknown and undulating terrain. These considerations require a robot to be adaptive in its immediate actions in order to successfully achieve long-term, resource-efficient and robust autonomy. To achieve this, we propose a hierarchical framework that integrates a local dynamic path planner with a longer term objective based planner and advanced motion control methods, whilst taking into consideration the dynamic responses of moving individuals within the environment. The framework is motivated by and synthesizes our recent work on energy aware mission planning, path planning in dynamic environments, and receding horizon motion control. In this paper we detail the proposed framework and outline its integration on a robotic platform. We evaluate the strategy in extensive simulated trials, traversing between objective waypoints to complete tasks such as soil sampling, weeding and recharging across a dynamic environment, demonstrating its capability to robustly adapt long term mission plans in the presence of moving individuals and obstacles for real world applications such as large scale farming.

Stuart Eiffert, Kunming Li, Mao Shan et al.

Understanding and predicting the intention of pedestrians is essential to enable autonomous vehicles and mobile robots to navigate crowds. This problem becomes increasingly complex when we consider the uncertainty and multimodality of pedestrian motion, as well as the implicit interactions between members of a crowd, including any response to a vehicle. Our approach, Probabilistic Crowd GAN, extends recent work in trajectory prediction, combining Recurrent Neural Networks (RNNs) with Mixture Density Networks (MDNs) to output probabilistic multimodal predictions, from which likely modal paths are found and used for adversarial training. We also propose the use of Graph Vehicle-Pedestrian Attention Network (GVAT), which models social interactions and allows input of a shared vehicle feature, showing that inclusion of this module leads to improved trajectory prediction both with and without the presence of a vehicle. Through evaluation on various datasets, we demonstrate improvements on the existing state of the art methods for trajectory prediction and illustrate how the true multimodal and uncertain nature of crowd interactions can be directly modelled.

Stuart Eiffert, He Kong, Navid Pirmarzdashti et al.

State of the art methods for robotic path planning in dynamic environments, such as crowds or traffic, rely on hand crafted motion models for agents. These models often do not reflect interactions of agents in real world scenarios. To overcome this limitation, this paper proposes an integrated path planning framework using generative Recurrent Neural Networks within a Monte Carlo Tree Search (MCTS). This approach uses a learnt model of social response to predict crowd dynamics during planning across the action space. This extends our recent work using generative RNNs to learn the relationship between planned robotic actions and the likely response of a crowd. We show that the proposed framework can considerably improve motion prediction accuracy during interactions, allowing more effective path planning. The performance of our method is compared in simulation with existing methods for collision avoidance in a crowd of pedestrians, demonstrating the ability to control future states of nearby individuals. We also conduct preliminary real world tests to validate the effectiveness of our method.

Stuart Eiffert, Salah Sukkarieh

Robotic navigation through crowds or herds requires the ability to both predict the future motion of nearby individuals and understand how these predictions might change in response to a robot's future action. State of the art trajectory prediction models using Recurrent Neural Networks (RNNs) do not currently account for a planned future action of a robot, and so cannot predict how an individual will move in response to a robot's planned path. We propose an approach that adapts RNNs to use a robot's next planned action as an input alongside the current position of nearby individuals. This allows the model to learn the response of individuals with regards to a robot's motion from real world observations. By linking a robot's actions to the response of those around it in training, we show that we are able to not only improve prediction accuracy in close range interactions, but also to predict the likely response of surrounding individuals to simulated actions. This allows the use of the model to simulate state transitions, without requiring any assumptions on agent interaction. We apply this model to varied datasets, including crowds of pedestrians interacting with vehicles and bicycles, and livestock interacting with a robotic vehicle.

Nathan Wallace, He Kong, Andrew Hill et al.

The control of field robots in varying and uncertain terrain conditions presents a challenge for autonomous navigation. Online estimation of the wheel-terrain slip characteristics is essential for generating the accurate control predictions necessary for tracking trajectories in off-road environments. Receding horizon estimation (RHE) provides a powerful framework for constrained estimation, and when combined with receding horizon control (RHC), yields an adaptive optimisation-based control method. Presently, such methods assume slip to be constant over the estimation horizon, while our proposed structured blocking approach relaxes this assumption, resulting in improved state and parameter estimation. We demonstrate and compare the performance of this method in simulation, and propose an overlapping-block strategy to ameliorate some of the limitations encountered in applying noise-blocking in a receding horizon estimation and control (RHEC) context.

Akash Arora, P. Michael Furlong, Robert Fitch et al.

Robotic science missions in remote environments, such as deep ocean and outer space, can involve studying phenomena that cannot directly be observed using on-board sensors but must be deduced by combining measurements of correlated variables with domain knowledge. Traditionally, in such missions, robots passively gather data along prescribed paths, while inference, path planning, and other high level decision making is largely performed by a supervisory science team. However, communication constraints hinder these processes, and hence the rate of scientific progress. This paper presents an active perception approach that aims to reduce robots' reliance on human supervision and improve science productivity by encoding scientists' domain knowledge and decision making process on-board. We use Bayesian networks to compactly model critical aspects of scientific knowledge while remaining robust to observation and modeling uncertainty. We then formulate path planning and sensor scheduling as an information gain maximization problem, and propose a sampling-based solution based on Monte Carlo tree search to plan informative sensing actions which exploit the knowledge encoded in the network. The computational complexity of our framework does not grow with the number of observations taken and allows long horizon planning in an anytime manner, making it highly applicable to field robotics. Simulation results show statistically significant performance improvements over baseline methods, and we validate the practicality of our approach through both hardware experiments and simulated experiments with field data gathered during the NASA Mojave Volatiles Prospector science expedition.

Akash Arora, Robert Fitch, Salah Sukkarieh

Autonomous Science is a field of study which aims to extend the autonomy of exploration robots from low level functionality, such as on-board perception and obstacle avoidance, to science autonomy, which allows scientists to specify missions at task level. This will enable more remote and extreme environments such as deep ocean and other planets to be studied, leading to significant science discoveries. This paper presents an approach to extend the high level autonomy of robots by enabling them to model and reason about scientific knowledge on-board. We achieve this by using Bayesian networks to encode scientific knowledge and adapting Monte Carlo Tree Search techniques to reason about the network and plan informative sensing actions. The resulting knowledge representation and reasoning framework is anytime, handles large state spaces and robust to uncertainty making it highly applicable to field robotics. We apply the approach to a Mars exploration mission in which the robot is required to plan paths and decide when to use its sensing modalities to study a scientific latent variable of interest. Extensive simulation results show that our approach has significant performance benefits over alternative methods. We also demonstrate the practicality of our approach in an analog Martian environment where our experimental rover, Continuum, plans and executes a science mission autonomously.