Victor C. Aitken

Multisensor vehicle-mounted teleoperated mine detector with data fusion

The Improved Landmine Detector Project (ILDP) was initiated in Autumn 1994 to develop a prototype teleoperated vehicle mounted mine detector for low metal content and nonmetallic mines to meet the Canadian requirements for rear area mine clearance in combat situations and peacekeeping on roads and tracks. The relatively relaxed requirements, such as low speed and reduced detectability of completely nonmetallic mines, greatly increase the likelihood of success. The ILDP system consists of a unique teleoperated vehicle carrying a forward looking infrared imager, a 3 m wide down-looking highly sensitive electromagnetic induction detector and a 3 m wide down-looking ground probing radar, which all scan the ground in front of the vehicle. Scanning sensor information is combined using a suite of navigation sensors and custom designed navigation, spatial correspondence and data fusion algorithms. Suspect targets are then confirmed by a thermal neutron analysis detector. A key element to the success of the system is the combination of sensor information. This requires coordinated communication between the sensors and navigation system and well designed sensor co-registration, spatial correspondence and data fusion methodologies. These complex tasks are discussed in detail. The advanced development model was completed in October 1997 and testing and improvements are ongoing. Results of system performance during extensive field trials are presented. A follow-on project has been initiated to build four to six production units for the Canadian Forces by the year 2000.

Evidential Mapping for Mobile Robots with Range Sensors

Mapping for mobile robots integrates noisy spurious sensor data into a single coherent map useful for navigational purposes. There are various frameworks used for mapping, but the Bayesian framework appears to be most popular. In this paper, the theory behind the Bayesian framework as it is used in mapping is briefly compared to a framework based on evidential theory. The remainder of this paper evaluates the use of the evidential framework by simulating its use on a mobile robot with sparse range sensors. A sensor model is described for the range sensors to work with evidential mapping, and the framework was evaluated under varying parameters and in different simulated test environments

The Auxiliary Extended and Auxiliary Unscented Kalman Particle Filters

This paper proposes two new particle filters, namely, the auxiliary extended Kalman particle filter (AEKPF) and the auxiliary unscented Kalman particle filter (AUKPF). The theory governing the newly proposed filtering techniques is developed and the algorithms are described and contrasted. Next, a series of tests is presented in which the new filters are compared against the extended Kalman filter (EKF), the unscented Kalman filter (UKF), and several existing particle filters. The test results are from simulations with synthetic mathematical models that incorporate elements that are nonlinear, non-stationary, and stochastic. Performance results are presented for various degrees of model nonlinearity including first, second, and third order systems. Furthermore, experimental results are also reported comparing the filters performances with different signal to noise ratios and noise models, including Gaussian, Cauchy, and Gamma distributions. Various metrics are used to compare the filters performances and to make conclusions about future work. It is shown to be advantageous to use certain particle filters depending on the noise distribution of the system of interest. In particular, the AUKPF and the AEKPF outperform existing particle filters in many cases.

Review of measurement quality metrics for range imaging

Quality metrics, within the field of laser range imaging, are used to quantify by how much some aspect of a measurement deviates from a predefined standard. Measurement quality evaluations are becoming increasingly important in laser range imaging for range image registration, merging measurements, and planning the next best view. Spatial uncertainty and resolution are the primary metrics of image quality; however, spatial uncertainty is affected by a variety of environmental factors. A review how contemporary researchers have attempted to quantify these environmental factors is presented, along with spatial uncertainty and resolution, resulting in a wide range of quality metrics.

Adaptive Laser Range Scanning using Quality Metrics

We present an approach to laser range scanning in which quality metrics are used to automatically reduce the number of measurements acquired from a scanner viewpoint in order to guide a minimally trained operator through the scanning process. As part of this approach we present improved versions of the orientation and reflectivity quality metrics, and introduce six new within-scan quality metrics: outlier, enclosed, resolvability, planarity, integration, and aliasing. These metrics are combined to generate a total within-scan quality metric for each measurement in the scan. The orientation, resolvability, reflectivity, and planarity quality metrics are used to divide the total field of view into regions based on their likelihood to produce useful measurements. A series of small high-density raster scans is then automatically generated to cover regions automatically identified as having a significant likelihood to produce useful measurements. All scans are then merged to generate a composite range image. The total number of measurements in the composite range image is minimized by merging statistically close measurements using a minimum variance estimator weighted by the total within-scan quality of each measurement.

Towards robust discrete-time sliding mode observers

A sliding mode observer for discrete-time linear systems is proposed which can maintain the sliding mode and retains robustness properties of the continuous-time counterpart. Matching conditions under which the sliding mode is invariant to dynamical modelling errors and disturbances, and effects of measurement noise are also presented.

A Comparison of Precision and Accuracy in Triangulation Laser Range Scanners

Precision and accuracy-based models for well-calibrated time-of-flight-based laser range scanners can be assumed to agree; however, this is not the case for triangulation-based laser range scanners. We show that the accuracy model for a triangulation laser range scanner features a skewed distribution even if the precision model is assumed to follow a Gaussian distribution. This is because the magnitude of the sensor uncertainty for any given measurement is a function of range which violates the assumption of identical distribution. We demonstrate, using a simulation, that this effect should be able to be detected using a properly-designed experiment. Future work will involve applying the test protocol to a real laser range scanner

Constraint consensus concentration for identifying disjoint feasible regions in nonlinear programmes

It is usually not known in advance whether a nonlinear set of constraints has zero, one, or multiple feasible regions. Further, if one or more feasible regions exist, their locations are usually unknown. We propose a method for exploring the variable space quickly, using constraint consensus (CC) to identify promising areas that may contain a feasible region. Multiple CC solution points are clustered to identify regions of attraction. A new inter-point distance frequency distribution technique is used to determine the critical distance for the single-linkage clustering algorithm, which in turn determines the estimated number of disjoint feasible regions. The effectiveness of multistart global optimization is increased due to better exploration of the variable space, and efficiency is also increased because the expensive local solver is launched just once near each identified feasible region. The method is demonstrated on a variety of highly nonlinear models.

Analysis and Comparison of the Generic and Auxiliary Particle Filtering Frameworks

State estimation is of paramount importance in many fields of engineering. Filtering is the method of estimating the state of a system by incorporating noisy observations as they become available online with prior knowledge of the system model. Particle filters are sequential Monte Carlo methods that use a point mass representation of probability densities in order to propagate the required statistical properties for state estimation. This paper is a quantitative comparison of the generic and auxiliary particle filtering frameworks using various proposal densities and state characterizations. New particle filtering methods that use the extended and unscented Kalman filters as state characterizations in the auxiliary framework are introduced. All the methods are compared in terms of accuracy and robustness. A synthetic stochastic model that incorporates nonlinear, non-stationary, and non-Gaussian elements is used for the experiments. It is shown that the particle filters designed with the auxiliary framework outperform the generic particle filters and other nonlinear filtering methods in this experiment

Uniform clustered particle filtering for robot localization

Localization is a fundamental ability for an autonomous mobile robot. Different particle filter based solutions to the problem are simulated in a software simulator, and the results are compared and discussed. The weighted bootstrap alter, clustering particle alter, and the uniform particle alter are examined. A new method based on the clustered particle alter and the uniform particle alter the uniform clustering particle alter, is also proposed and evaluated.