Dealing safely with nuclear waste is an imperative for the nuclear industry. Increasingly, robots are being developed to carry out complex tasks such as perceiving, grasping, cutting, and manipulating waste. Radioactive material can be sorted, and either stored safely or disposed of appropriately, entirely through the actions of remotely controlled robots. Radiological characterisation is also critical during the decommissioning of nuclear facilities. It involves the detection and labelling of radiation levels, waste materials, and contaminants, as well as determining other related parameters (e.g., thermal and chemical), with the data visualised as 3D scene models. This paper overviews work by researchers at the QMUL Centre for Advanced Robotics (ARQ), a partner in the UK EPSRC National Centre for Nuclear Robotics (NCNR), a consortium working on the development of radiation-hardened robots fit to handle nuclear waste. Three areas of nuclear-related research are covered here: human–robot interfaces for remote operations, sensor delivery, and intelligent robotic manipulation.

This paper addresses the problem of range-stereo fusion, for the construction of high-resolution depth maps. In particular, we combine low-resolution depth data with high-resolution stereo data, in a maximum a posteriori (MAP) formulation. Unlike existing schemes that build on MRF optimizers, we infer the disparity map from a series of local energy minimization problems that are solved hierarchically, by growing sparse initial disparities obtained from the depth data. The accuracy of the method is not compromised, owing to three properties of the data-term in the energy function. First, it incorporates a new correlation function that is capable of providing refined correlations and disparities, via subpixel correction. Second, the correlation scores rely on an adaptive cost aggregation step, based on the depth data. Third, the stereo and depth likelihoods are adaptively fused, based on the scene texture and camera geometry. These properties lead to a more selective growing process which, unlike previous seed-growing methods, avoids the tendency to propagate incorrect disparities. The proposed method gives rise to an intrinsically efficient algorithm, which runs at 3FPS on 2.0 MP images on a standard desktop computer. The strong performance of the new method is established both by quantitative comparisons with state-of-the-art methods, and by qualitative comparisons using real depth-stereo data-sets.

Robotic grasping is fundamental to many real-world applications, and new approaches must be systematically evaluated. However, in most cases, the performance of a specific approach is assessed by simply counting the number of successful attempts in a given task, and this success rate is then compared to those of other solutions, without taking into account the random variability across different experiments (e.g. due to sensor noise or variations in object placement). In order to address this issue, we classify the observed performance into qualitatively ordered outcomes, thereby stratifying the results. We then show how to analyze these results in a statistical framework, which accounts for the variability between experiments. The advantages of our approach are demonstrated in the practical comparison of four grasp planning algorithms. In particular, we show that the proposed approach allows us to carry out several distinct evaluations from a single set of experiments, without having to repeat the data collection process. We demonstrate that differences between the algorithms, which would not be apparent from overall success rates, can be identified and evaluated.

Accurate object detection and location systems are essential for many robotic applications, including autonomous grasping and manipulation systems. In some cases, the target object may be obscured from view, in clutter, packaging, or debris. Millimetre wave radar is a potential alternative to visual sensing in such scenarios, owing to its ability to penetrate typical low-density non-metallic materials. However, this approach requires accurate spatial calibration of the radar signal, over the robot workspace. We propose to achieve this with reference to visual data, which provides ground truth locations for initial training of the system. Specifically, we describe a commodity millimetre wave radar system for detecting and localizing static metallic objects, over a 2D workspace. We compare similarity, affine, and thin-plate spline models of the spatial transformation from radar estimates to actual locations. Experiments were performed with a frequency modulated continuous wave (FMCW)multiple-inputmultiple-output (MIMO) device, using a starting frequency of 60 GHz and a bandwidth of 3.4 GHz. It is shown that the spline model performs best, achieving an average spatial error of 7 mm, which is an order of magnitude lower than that of the uncalibrated system.