An autonomous vehicle would benefit from being able to predict trajectories of other vehicles in its vicinity for improved safety. In order for the self-driving car to plan safe trajectories, paths of nearby vehicles are required to be predicted for risk assessment, decision making, and motion planning. In this study, a trajectory prediction algorithm based on the Interacting Multiple Model (IMM) estimation strategy is proposed to predict paths involving lane-changing, lane-keeping, and turning motion. More specifically, the Interacting Multiple Model estimation technique is used with models defined in curvi-linear coordinates to predict a vehicle’s trajectory based on prior behavioral maneuvers. The road geometry is used to help facilitate behavior identification and prediction. Moreover, the combination of a more recently developed estimation technique known as the Generalized Variable Boundary Layer-Smooth Variable Structure Filter and the Interacting Multiple Model Estimator is applied to track, identify behaviors, and predict trajectories of a vehicle. The performance of this technique is compared with a Kalman Filter based formulation using synthetic and experimental data. This model-based strategy is also compared with machine learning-based strategies for trajectory prediction.

This work is a study of the current state of vehicle trajectory analysis in intersections monitored by a single fisheye camera. The challenges related to using fisheye cameras at intersections are explained, and a zone definition-based framework to detect illegal maneuvers is introduced for busy intersections containing noisy detection and tracking data. The methodology is developed by conducting data rectification, distance-based data filtering, and trajectory classification, followed by detecting banned maneuvers. The illegal maneuver detection algorithm is further extended for defining different banned maneuvers, including illegal turns and box-blocking contravention at intersections. Lastly, the proposed framework is implemented in real-world data captured from the Cubic’s GRIDSMART bell camera, and the results are explained in detail.

Autonomous vehicles need a real-time traffic tracking system in order to interact with multiple moving vehicles in urban situations. Most existing works rely on camera sensors to interpret the surrounding environment. However, camera sensors degrade in conditions of lighting, shadows, and extreme weathers such that they can hardly detect objects. The light detection and ranging (LiDAR) sensors promise to be a good substitute, as they enable highly precise and robust localization across a wide range of conditions. This paper presents a new LiDAR-based online multiple maneuvering vehicle tracking problem and proposes a novel online multi-model smooth variable structure filter to address the problem. The real-time experiments show that our method is able to deliver superior performance compared to other conventional methods.

Autonomous vehicles have been emerging over the past few years because of the sophisticated processing of data from different types of perception sensors, such as LiDAR, radar, and camera. Ground segmentation plays an important role in the sequence of data processing for environment perception tasks, as it can help to reduce the size of data to be processed and further decrease the overall computational time. However, the over-segmentation, under-segmentation, or slow-segmentation on non-flat surface usually occurs due to the characteristics of the 3D LIDAR data, such as occlusion in complex urban environment. To address these problems, in this paper, we proposed a probability occupancy grid-based approach for real-time ground segmentation by using a single LiDAR sensor. The effectiveness and robustness of our proposed method are evaluated and demonstrated by the real-time experiments that span different traffic scenarios from heavy traffic to light traffic.

Autonomous self-drive requires intelligence and cognition that relies on observations and tracking of the state of motion of surrounding vehicles. This information can be acquired by using sensors; however, these are often affected by clutter and noise that, in turn, introduce the issues of estimation and data origin uncertainty into the tracking system. The most popular methods for estimation and tracking are based on the well-studied Kalman filter (KF). KF is optimal when noise is white and remains so despite uncertainties in the filter model; the robustness and stability of the KF is affected if this condition is not met. The smooth variable structure filter (SVSF) is a relatively new method that is more robust to disturbances and uncertainties. The SVSF ensures stability by using a discontinuous corrective term that maintains estimates to within a subspace of the true state trajectory. The discontinuous corrective term results in chattering that is removed by using a smoothing boundary layer. In this paper, a generalized covariance formulation of the SVSF and a generalized optimal time-varying smoothing boundary layer are proposed. The generalized optimal SVSF is then combined with a joint probabilistic data association technique for target tracking. The robustness and accuracy of the new form of filtering and data association is validated and comparatively analyzed by its application to an experimental traffic monitoring system based on Light Detection and Ranging.

In this paper, we describe an adaptive technique for states and parameter estimation involving a combination of two methods, namely the Variable Structure Filter (VSF) and the Extend Kalman Filters (EKF).

In this paper, a new method for state estimation, referred to as the smooth variable structure filter (SVSF), is presented. The SVSF method is model based and applies to smooth nonlinear dynamic systems. It allows for the explicit definition of the source of uncertainty and can guarantee stability given an upper bound for uncertainties and noise levels. The performance of the SVSF improves with more refined definition of upper bounds on parameter variations or uncertainties. Furthermore, most filtering methods provide as their measure of performance the filter innovation vector or (output) estimation error. However in addition to the innovation vector, the SVSF has a secondary set of performance indicators that correlate to the modeling errors specific to each state or parameter that is being estimated. The combined robustness and multiple indicators of performance allow for dynamic refinement of internal models in the SVSF. Dynamic refinement and robustness are features that are particularly advantageous in fault diagnosis and prediction. In this paper, the applications of the SVSF to linear and nonlinear systems, including one pertaining to fault detection, are provided. The characteristics of this filter in terms of its accuracy and rate of convergence are discussed.

Since the advent of robotics, manipulators have, in practice, been confined to specific application areas. This limitation can be attributed to their geometrical and physical characteristics such as having fixed link lengths, fixed angular ranges, relatively large size and mass, and their torque capabilities. This paper describes the design of a heavy-duty hydraulic manipulator, the Guelph Arm, which is uniquely flexible in its applications, The features of this robot include compactness, reconfigurability, variable link lengths, large torque and low, overall mass. The Guelph Arm is modeled and, the design and implementation of an industrial controller is given. The use of this robot in neurophysiological experiments is reported.