Robotic agents are now ubiquitous in both home and work environments; moreover, the degree of task complexity they can undertake is also increasing exponentially. Now that advanced robotic agents are commonplace, the question for utilisation becomes how to enable collaboration of these agents, and indeed, many have considered this over the last decade. If we can leverage the heterogeneous capabilities of multiple agents, not only can we achieve more complex tasks, but we can better position the agents in more chaotic environments and compensate for lacking systems in less sophisticated agents. Environments such as search and rescue, agriculture, autonomous vehicles, and robotic maintenance are just a few examples of complex domains that can leverage collaborative robotics. If the use of a robotic agent is fruitful, the question should be: How can we provide a world state and environment mapping, combined with a communication method, that will allow these robotic agents to freely communicate? Moreover, how can this be decentralised such that agents can be introduced to new and existing environments already understood by other agents? The key problem that is faced is the communication method; however, when looking deeper we also need to consider how the change of an environment is mapped while considering that there are multiple differing sensors. To this end, we present the voxel grid approach for use in a decentralised robotic colony. To validate this, results are presented to show how the single-agent and multiagent systems compare.

Extended objects generate a variable number of multiple measurements. In contrast with point targets, extended objects are characterized with their size or volume, and orientation. Multiple object tracking is a notoriously challenging problem due to complexities caused by data association. This paper develops a box particle filter (box PF) method for multiple extended object tracking, and for the first time, it is shown how interval-based approaches can deal efficiently with data association problems and reduce the computational complexity of the data association. The box PF relies on the concept of a box particle. A box particle represents a random sample and occupies a controllable rectangular region of nonzero volume in the object state space. A theoretical proof of the generalized likelihood of the box PF for multiple extended objects is given based on a binomial expansion. Next, the performance of the box PF is evaluated using a challenging experiment with the appearance and disappearance of objects within the area of interest, with real laser rangefinder data. The box PF is compared with a state-of-the-art particle filter with point particles. Accurate and robust estimates are obtained with the box PF, both for the kinematic states and extent parameters, with significant reductions in computational complexity. The box PF reduction of computational time is atleast 32% compared with the particle filter working with point particles for the experiment presented. Another advantage of the box PF is its robustness to initialization uncertainty.

In this paper, we review the five rules published in EPSRC Principles of Robotics with a specific focus on future robotics research topics. It is demonstrated through a pictorial representation of the five rules that these rules are questionably not sufficient, overlapping and not explicitly reflecting the true challenges of robotics ethics in relation to the future of robotics research.

Autonomous systems such as Unmanned Aerial Vehicles (UAVs) need to be able to recognise and track crowds of people, e.g. for rescuing and surveillance purposes. Large groups generate multiple measurements with uncertain origin. Additionally, often the sensor noise characteristics are unknown but measurements are bounded within certain intervals. In this work we propose two solutions to the crowds tracking problem— with a box particle filtering approach and with a convolution particle filtering approach. The developed filters can cope with the measurement origin uncertainty in an elegant way, i.e. resolve the data association problem. For the box particle filter (PF) we derive a theoretical expression of the generalised likelihood function in the presence of clutter. An adaptive convolution particle filter (CPF) is also developed and the performance of the two filters is compared with the standard sequential importance resampling (SIR) PF. The pros and cons of the two filters are illustrated over a realistic scenario (representing a crowd motion in a stadium) for a large crowd of pedestrians. Accurate estimation results are achieved.

This paper considers the problem of tracking a target - which might or might not exist - from a platform whose position is not known perfectly and might contain substantial time dependencies.Most single and multi-target tracking algorithms implicitly or explicitly assume that the location of the sensor platform system is known perfectly. However, in practice the location of sensing platforms is often estimated, usually by fusing a set of sensor measurements from different sources. As a result, the error in the platform estimates could be significant and time correlated. These difficulties are compounded in single and multi-target tracking problems when the existence of a target is not guaranteed.In this paper, we consider the problem of tracking at most a single target from a poorly-localized UAV. We develop a formulation of the Bernoulli filter which incorporates both the target state and the state of the platform. However, because the dimension of the state is relatively large, we develop a suboptimal algorithm which, through neglecting the use of track information to improve the quality of the platform estimate, scales in a manner very similar to that of a conventional Bernoulli filter.The implementations of the different algorithms are tested in a simulation scenario of a UAV performing safety monitoring of a convoy.

This paper focuses on tracking large groups of objects, such as crowds of pedestrians. Large groups generate multiple measurements with uncertain origin. Additionally, often the sensor noise characteristics are unknown but bounded within known intervals. Hence, these two types of uncertainties call for flexible techniques capable of offering a solution in the presence of data association and also to cope with the presence of nonlinearities. This paper presents a box particle filter for large crowds tracking able to deal with such challenges. The filter measurement update step is performed by solving a dynamic constraint satisfaction problem (DSCP) with the multiple measurements. The box particle filter performance is validated over a realistic scenario comprising a large crowd of pedestrians. Promising results are presented in terms of accuracy and computational complexity.

Multitarget tracking is fundamental in many security and surveillance applications. However, as algorithms have been applied in more and more challenging environments, traditional simplifications no longer even approximately hold true. In particular, we consider the problem of maritime surveillance in which small targets (boats) are to be detected and tracked in the presence of highly structured noise (waves). In particular, we consider two problems. The first is that, given the resolution of modern radar systems, even small targets are extended, straddling multiple range or azimuth bins. The second is that sea clutter is not a uniform, Poisson-distributed noise process but is highly spatially varying. In this paper, we develop two extensions of the Probability Hypothesis Density (PHD) Filter. Using a generalised likelihood model, extended targets can be readily accounted for. Through the use of spatially varying clutter models, structured noise approximation is provided. The algorithms were developed and tested using a sea trial in which Rigid-Hulled Inflatable Boats (RHIBs), equipped with GPS receivers, were tracked using a radar system.

This paper considers the problem of localization of multiple land mines using information collected by a network of wireless sensors deployed in the area of interest. These sensors detect the concentration of explosive vapours, emanating from buried land mines, in the air and have a wireless transmission system for exchanging information with a treatment or fusion centre. One of the key contribution of this paper is to locate and estimate the emission rate of multiple land mines. Using a model for the transport of the explosive chemicals in the air, we formulate the inverse problem consisting in determining sequentially the positions and emission rates of the land mines knowing concentration measurements provided by the sensors. To solve the inverse problem, we present a first solution based on a Least Squares optimization approach and a second solution based on probabilistic Bayesian techniques using a Markov Chain Monte Carlo sampling scheme. These two approaches are tested and compared on simulated data.

. Extended objects generate multiple measurements and are characterised with their size or volume. They require methods able to deal with the data association problem and at the same time to estimate both their kinematic states and shape parameters. This paper presents a solution to the extended object tracking for rectangular extended objects, with the Box Particle filter (Box PF) approach. The Box PF is implemented based on dynamically calculated constraints. Promising results are demonstrated.

We consider the problem of localising an unknown number of land mines usingconcentration information provided by a wireless sensor network. A number of vapoursensors/detectors, deployed in the region of interest, are able to detect the concentrationof the explosive vapours, emanating from buried land mines. The collected data iscommunicated to a fusion centre. Using a model for the transport of the explosive chemicalsin the air, we determine the unknown number of sources using a Principal ComponentAnalysis (PCA)-based technique. We also formulate the inverse problem of determiningthe positions and emission rates of the land mines using concentration measurementsprovided by the wireless sensor network. We present a solution for this problem basedon a probabilistic Bayesian technique using a Markov chain Monte Carlo sampling scheme,and we compare it to the least squares optimisation approach. Experiments conducted onsimulated data show the effectiveness of the proposed approach.

Through its ability to create situation awareness, multi-target target tracking is an extremely important capability for almost any kind of surveillance and tracking system. Many approaches have been proposed to address its inherent challenges. However, the majority of these approaches make two assumptions: the probability of detection and the clutter rate are constant. However, neither are likely to be true in practice. For example, as the projected size of a target becomes smaller as it moves further from the sensor, the probability of detection will decline. When target detection is carried out using templates, clutter rate will depend on how much the environment resembles the current target of interest. In this paper, we begin to investigate the impacts on these effects. Using a simulation environment inspired by the challenges of Wide Area Surveillance (WAS), we develop a state dependent formulation for probability of detection and clutter. The impacts of these models are compared in a simulated urban environment populated by multiple vehicles and cursed with occlusions. The results show that accurate modelling the effects of occlusion and degradation in detection, significant improvements in performance can be obtained.

This paper deals with the problem of inference in distributed systems where the probability model is stored in a distributed fashion. Graphical models provide powerful tools for modeling this kind of problems. Inspired by the box particle filter which combines interval analysis with particle filtering to solve temporal inference problems, this paper introduces a belief propagation-like message-passing algorithm that uses bounded error methods to solve the inference problem defined on an arbitrary graphical model. We show the theoretic derivation of the novel algorithm and we test its performance on the problem of calibration in wireless sensor networks. That is the positioning of a number of randomly deployed sensors, according to some reference defined by a set of anchor nodes for which the positions are known a priori. The new algorithm, while achieving a better or similar performance, offers impressive reduction of the information circulating in the network and the needed computation times.

This paper develops a novel approach for multitarget tracking, called box-particle probability hypothesis density filter (box-PHD filter). The approach is able to track multiple targets and estimates the unknown number of targets. Furthermore, it is capable of dealing with three sources of uncertainty: stochastic, set-theoretic, and data association uncertainty. The box-PHD filter reduces the number of particles significantly, which improves the runtime considerably. The small number of box-particles makes this approach attractive for distributed inference, especially when particles have to be shared over networks. A box-particle is a random sample that occupies a small and controllable rectangular region of non-zero volume. Manipulation of boxes utilizes methods from the field of interval analysis. The theoretical derivation of the box-PHD filter is presented followed by a comparative analysis with a standard sequential Monte Carlo (SMC) version of the PHD filter. To measure the performance objectively three measures are used: inclusion, volume, and the optimum subpattern assignment (OSPA) metric. Our studies suggest that the box-PHD filter reaches similar accuracy results, like an SMC-PHD filter but with considerably less computational costs. Furthermore, we can show that in the presence of strongly biased measurement the box-PHD filter even outperforms the classical SMC-PHD filter.

Resulting from the synergy between the sequential Monte Carlo (SMC) method [1] and interval analysis [2], box particle filtering is an approach that has recently emerged [3] and is aimed at solving a general class of nonlinear filtering problems. This approach is particularly appealing in practical situations involving imprecise stochastic measurements that result in very broad posterior densities. It relies on the concept of a box particle that occupies a small and controllable rectangular region having a nonzero volume in the state space. Key advantages of the box particle filter (box-PF) against the standard particle filter (PF) are its reduced computational complexity and its suitability for distributed filtering. Indeed, in some applications where the sampling importance resampling (SIR) PF may require thousands of particles to achieve accurate and reliable performance, the box-PF can reach the same level of accuracy with just a few dozen box particles. Recent developments [4] also show that a box-PF can be interpreted as a Bayes? filter approximation allowing the application of box-PF to challenging target tracking problems [5].

This paper presents a novel method for solving nonlinear filtering problems. This approach is particularly appealing in practical situations involving imprecise stochastic measurements, thus resulting in very broad posterior densities. It relies on the concept of a box particle, which occupies a small and controllable rectangular region having a non-zero volume in the state space. Key advantages of the box particle filter (Box-PF) against the standard particle filter (PF) are in its reduced computational complexity and its suitability for distributed filtering. Indeed, in some applications where the sequential importance resampling (SIR) PF may require thousands of particles to achieve an accurate and reliable performance, the Box-PF can reach the same level of accuracy with just a few dozens of box particles.

In state estimation theory, the general formulation is often done under assumptions of stochastic noise processes obeying well known probability distributions such as the Gaussian family. However, in many practical applications, due to the presence of high non-linearities and unknown noise probability distributions, other methods are required. Methods such as imprecise probabilities and set-membership approaches offer robust alternative solutions to the lack of statistical information. In these frameworks, the solution to the estimation problem is no longer a posterior distribution but either a set of densities or a solution set in the state space. The main objective in this work is to take advantage of both Monte Carlo approaches and set membership methods. A novel approach to non-linear non-Gaussian state estimation problems is presented based on mixtures of imprecise samples which can be seen as unknown probability density functions with known supports. The derivation of a sequential Bayesian procedure and convergence properties of such a representation are provided.

This work presents the current state-of-the-art in techniques for tracking a number of objects moving in a coordinated and interacting fashion. Groups are structured objects characterized with particular motion patterns. The group can be comprised of a small number of interacting objects (e.g. pedestrians, sport players, convoy of cars) or of hundreds or thousands of components such as crowds of people. The group object tracking is closely linked with extended object tracking but at the same time has particular features which differentiate it from extended objects. Extended objects, such as in maritime surveillance, are characterized by their kinematic states and their size or volume. Both group and extended objects give rise to a varying number of measurements and require trajectory maintenance. An emphasis is given here to sequential Monte Carlo (SMC) methods and their variants. Methods for small groups and for large groups are presented, including Markov Chain Monte Carlo (MCMC) methods, the random matrices approach and Random Finite Set Statistics methods. Efficient real-time implementations are discussed which are able to deal with the high dimensionality and provide high accuracy. Future trends and avenues are traced.

We present an inference and data fusion method for tracking maritime Fast Inshore Attack Craft (FIACs) using multiple sensors. The scenario addressed encompasses littoral, counter-piracy and maritime constabulary operations. The problem space is characterised by mixed sensor modalities, non-stationary and spatially-varying non-Gaussian clutter, intermittent observations and a high false alarm rate. Our method combines the Probability Hypothesis Density (PHD) lter for multi-target Bayesian inference with Generalised Covariance Intersection (GCI) for decentralised data fusion. We outline the development and testing of our solution using electro-optical and radar observations of marine trac in the Solent. These data are complemented by ground truth positional data of marine trac including high-frequency positional estimates of two representative FIACs. Our system has been deployed both oine and in real time. We carry out a number of experiments designed to show the ecacy of the algorithms in representative scenarios. The performance of our algorithms is quantied using multi-target inference metrics. We show that the combination of PHD and GCI has many advantages over traditional inference and fusion methods, particularly in cluttered environments.