Graham A. Parker

Augmented reality as a tool to aid the telerobotic exploration and characterization of remote environments

The three-dimensional characterization and mapping of remote environments is an important task that generates a good deal of attention both by end users and by researchers across several fields of interest. In the mobile robotics community, a great deal of work has been done in equipping vehicles with sensors that can acquire three-dimensional and even multimodal information about the location and nature of features and objects in remote environments. However, the interpretation of such data using fully autonomous methods, such as computer vision, is usually a highly complex problem that, we believe, is much better suited to a humanoriented solution. In this paper, we describe our work in the development of augmented reality (AR) techniques for the telerobotic inspection and characterization of remote environments. We describe how we are using stereoscopic camera feedback from a remote vehicle and equipping the human operator with three-dimensional virtual cursors that can be used to interactively measure and model real features and objects in the remote environment. We include a description of the calibration techniques used to correctly align the real and virtual images both statically and under vehicle and camera motion. We also describe how we are using our system to demonstrate the potential of AR for improving the inspection of underground sewer pipes.

Some aspects of road sweeping vehicle automation

This paper reports on work being undertaken to enhance the performance and safety of road sweeping vehicles, including a forward looking debris detection system alongside research into brushes and the brushing process. Vision processing techniques are used to identify various road surfaces and debris. A continuous flow of data is provided by laser striping building up a 3D profile of the road ahead. Only when the scene cannot be fully identified with sizes and shapes alone, other more advanced image processing techniques are applied. A brush rig was constructed to test different brush types used on road sweeping vehicles. An extensive range of practical tests were undertaken on a variety of contact surfaces. Analysis of the results shows that the common brush types currently on the market have very different characteristics depending on the orientation of their tines.

Automatic detection of defects in industrial ultrasound images using a neural network

Time-of-flight diffraction (TOFD) is a relatively new method of ultrasonic inspection and is well suited to semi- automation using methods such as robotic scanning, computer conditioned data acquisition and signal and image enhancement. However very little work has been documented on the full computer understanding of such scans. Instead, most work has been directed at aiding the manual interpretation process to determine defect characteristics. This paper describes the application of image processing and neural networks (ANNs) to the task of completely automating the decision making process involved in the interpretation of TOFD scans. Local area analysis is used to derive a feature vector which contains 2D information on defect/component and non-defect areas. These vectors are then classified using an ANN trained with the backpropagation algorithm. The labelled image is then further segmented using binary shape analysis to discriminate between component echoes, or defect signals. Time-of-flight correction techniques may be then used in order to determine the location of defects within a scanned weld.

Brush Dynamics: Models and Characteristics

This paper reviews investigations into the dynamics and modelling of brushes. They include brushes for surface finishing operations, removal of fouling, post-CMP brushing processes, air duct cleaning, and street sweeping. The methods that have been proposed to model brush dynamics are described, and the results of the research into brush mechanics are presented and discussed. Some conclusions of the paper are as follows: brush dynamics is very complex, as it depends on the interaction among many phenomena and variables. The bristle oscillations that occur in some brushes constitute a complexity for modelling brush behaviour and are not normally addressed. Additionally, the literature reveals that the coefficient of friction is not a constant value that depends only on the materials and surface roughness of the two contacting bodies. Frictional behaviour strongly depends on many variables, such as brush setup angles and rotational speed, which play a part in the development of stick-slip friction cycles. Finally, it is concluded that brush behaviour and the phenomena involved in brushing have not been fully studied or understood and more research into this field is needed.Copyright

Teaching for multi-fingered robots based on motion intention in virtual reality

We present teaching for multi-fingered robots based on motion intention in virtual reality. In motion intention analysis, motion is divided into plural primitive motions, those unnecessary for executing the task are deleted, and the remainder are represented by a smooth time function. Segmentation is made using 3D motion measurement of human and virtual objects and virtual reaction generated by the hand of an operator using a force-feedback glove. Analyzed motion is represented by human motion commands. A robot teaching command for the object coordinate frame is generated from a human motion command. Experimental results of a pick-and-place task verify the feasibility of proposed robot teaching.

The Surrey attentive robot vision system

This paper presents the design and development of a real-time eye-in-hand stereo-vision system to aid robot guidance in a manufacturing environment. The stereo vision head comprises a novel camera arrangement with servo-vergence, focus, and aperture that continuously provides high-quality images to a dedicated image processing system and parallel processing array. The stereo head has four degrees of freedom but it relies on the robot end-effector for all remaining movement. This provides the robot with exploratory sensing abilities allowing it to undertake a wider variety of less constrained tasks. Unlike other stereo vision research heads, the overriding factor in the Surrey head has been a truly integrated engineering approach in an attempt to solve an extremely complex problem. The head is low cost, low weight, employs state-of-the-art motor technology, is highly controllable and occupies a small-sized envelope. Its intended applications include high-accuracy metrology, 3-D path following, object recognition and tracking, parts manipulation and component inspection for the manufacturing industry.

Experimental determination of optimum gutter brush parameters and road sweeping criteria for different types of waste

The removal ability of gutter brushes for road sweeping for various debris types and different sweeping parameters is studied through experimental tests. The brushing test rig used comprises two commercial gutter brushes, a concrete test bed, and an asphalt test road with a gutter of 0.25 cm width and 10° slope. The brush-surface contact area is determined by sweeping sand on the concrete test bed. Sweeping problems are identified and discussed, and sweeping criteria for the different debris types are suggested. Also, optimum sweeping parameters are proposed for each debris type. In addition, debris removal mechanisms are discussed and analysed. The results indicate that for large heavy debris such as stones and gravel, it is not difficult to achieve large removal forces, because the steel bristles are relatively stiff. Conversely, high removal forces are not needed for particles of millimetre or micron sizes, but bristle curvature has to be appropriate to remove particles from road concavities. Finally, it is found that mud, especially dry mud on a rough surface, is the hardest debris to sweep, requiring a brush with a large tilt angle and a very large penetration to produce large removal forces.

Brief paper: Practical nonlinear system identification using a modified volterra series approach

A functional approach has been developed to represent continuous separable, nonlinear systems of a general type based on a modified Volterra series. The importance of this is that the effects of bias or mean signal level within the nonlinear system can be separated from dynamic effects. This has particular significance in the development of identification procedures based on cross-correlation functions, as these functions can now be estimated practically without any influence from the bias level. Practical identification of the gain characteristics for an electrohydraulic servo is described using three-level pseudorandom input signals which are cross-correlated with the sampled system response in a minicomputer to provide an automated procedure. Good accuracy is achieved even in the presence of severe noise.

Effectiveness of gutter brushes in removing street sweeping waste

Litter on roadways has to be removed for hygiene and to reduce pollution, amongst other reasons. Therefore, the effective operation of street sweepers is important in the collection of solid waste. In this article, the effectiveness of gutter brushes of street sweepers in removing different debris types, namely medium-size gravel, small and fine particles, and wet thin debris is studied by means of sweeping tests using a brushing test rig. Two types of gutter brushes, cutting and F128, are tested under a variety of operating conditions. The experimental tests provide a means of identifying suitable ranges of brush penetration for the different debris types under defined operating parameters such as brush angle of attack, brush rotational speed, and sweeper velocity. These ranges may provide sufficiently high removal forces and avoid the occurrence of sweeping problems such as backward sweeping and inappropriate bristle-surface contact. Optimum operating parameters for the three debris types studied, as well as a mixture of all of them, are determined. Lastly, the results indicate that, as far as the sweeping effectiveness is concerned, the F128 brush is the preferred one for the cases studied.

Theoretical model for the dynamics of an unconstrained cutting brush of a street sweeper

A theoretical model for the free-flight behaviour of an oscillatory cutting brush-of a street sweeper is developed. The bristles are modelled as cantilever beams, and the equation of motion for the transverse vibrations is derived based on the theory of vibrations and small deflection beam theory. Two angular velocity functions are studied: a sinusoidal function and a function that provides small shaft accelerations and whose exact shape depends on a parameter b. The model is applied for a range of frequencies of oscillation that contains the first and second natural frequency of the bristle. The effects of the alternating component of the rotational speed, the type of function, and the value of b are also studied. The results are compared with those obtained in a previous work for a flicking brush. The findings suggest that the bending moment and bristle deflection tend to depend fairly linearly on the altemating velocity. In contrast to the bristles of a flicking brush, in the cutting brush resonance tends to occur only near the natural frequencies of the bristle. Additionally, the behaviour of the cutting brush is similar for both functions and for different values of b.