This paper presents the development and validation of an automated steering system that has been successfully implemented on a 18.3-m articulated bus for revenue service in Eugene, OR, USA. The automated steering system provides both lane keeping and S-curve precision docking on a 4-km-long narrow and curving urban segment with six stations and mixed traffic lanes while the operator controls the speed. The objective of deployment with an extended period of revenue service (carrying passengers) elevated the system safety and performance requirements. To meet these requirements, the controller needs to address accuracy, robustness, comfort, and fault tolerant capabilities, while the safety design needs to address functional safety of the automated system. The baseline high-gain controller designed for the automated bus adopts the steering mechanism of human drivers with optimized parameters. The fault tolerant control incorporates a family of Kalman filters to sustain the automation for a few seconds after part of the sensing system fails. The safety design followed the functional safety procedure and employed the combination of redundancy, multilayer fault detection, and comprehensive fault management. The resultant system achieved all performance requirements and the revenue service started in June 2013 and ended in May 2015.

This paper presents the design and development of an automated steering control system that has been successfully deployed on an 18.3-m (60-ft) articulated bus for revenue service in a bus rapid transit line in Eugene, OR, USA, between June 2013 and May 2015. This automated steering system provides both lane keeping and S-curve precision docking on 4-km-long narrow and curving urban segments with six stations and mixed traffic lanes while the operator controls the speed. The objective of deploying an automated bus that carries passengers for an extensive period of time elevated the system safety and performance requirements. This paper describes three key elements in the controller design that address the safety and performance challenges, i.e., high precision, fault tolerance, and control transitions. The experience of designing and deploying such a system in revenue service helps advance the field of automated vehicle by directly addressing its safety issues as an integral part of the automated vehicle control system design.

This paper focuses on the design of an automatic steering controller that has been implemented on an 18.3-m articulated bus for revenue service. Compared with the typical environment for autonomous cars, the narrow and curving bus lane and the tight S-shape docking curves require a steering controller to consistently provide high accuracy with ride comfort. While analyzing the data of drivers steering through a double lane change course, we discovered and validated that driver's steering rate is proportional to a specific vehicle angle error. Subsequent analysis showed that drivers in effect execute a naturally robust controller that allows high-gain corrections and is insensitive to variations in vehicle dynamics and speeds. This paper presents the first adaptation of this driver's steering mechanism to a real-world challenging application. The resultant system achieved all performance requirements, and the revenue service at Eugene, OR, USA, started in June 2013. Moreover, the simple control structure greatly facilitated the design of fault detection, degraded control, and control redundancy.

This paper describes the development of a driver steering model that captures driver key steering mechanisms based on the analyses of vehicle test data on the standard double lane change (DLC) course. These analyses indicate that, instead of planning and following a desired path according to the traditional trajectory-planning concept, drivers simply use the next lane center as the target points for control during lane changes. The data also suggest that drivers engage steering rate control instead of the conventional steering angle control to steer the vehicle. Accordingly, this paper proposes a relatively straightforward driver steering model based on this target and control scheme. Both the open-loop identification and closed-loop simulations verify that this relatively simple driver steering model is capable of capturing individual driver steering behavior and that the simulated steering rate matches well with the actual steering rate for all 80 vehicle test runs conducted by 20 different drivers.

This paper focuses on the robustness and reliability of a future-trajectory-based cooperative collision warning system (CCWS), which estimates and communicates vehicle positions and predicts and processes future trajectories for collision decision making. The feasibility of the system design has been demonstrated in previous work, and this paper analyzes the error propagation and its impact on detection performance. The reliability of the CCWS heavily depends on the reliability of the differential Global Positioning System (DGPS)-based positioning system and the reliability of intervehicle communication. This paper uses the Kalman filter technique to statistically characterize the errors in position estimation and trajectory prediction, incorporates communication errors as part of the prediction error, and accordingly determines the probability of conflicts and the quality of the detection performance. Results with experimental data are presented to illustrate the error propagation and to demonstrate the reliability of the system.

This paper describes the development process and the initial field test results of an automated snowblower, focusing on one of the more difficult snow removal operations: blowing snow off the freeway alongside a guardrail without the snowblower touching the guardrail. The development process includes transforming this highway winter maintenance operation into a control problem, modeling a snowblower, designing control algorithms, devising a human-machine interface, and equipping a 20-ton snowblower with sensors and an actuator. Specific challenges include modeling the low-speed tire-induced oscillation, designing high-gain automatic control on front wheels while keeping rear steering under driver control, and implementing such a system under practical limitations. A new dynamic deflection tire model is incorporated into a bicycle model to account for the additional lateral dynamics. A low-order controller was first generated based on the understanding of the specific control problem and, then, refined and tuned iteratively using linear-matrix-inequality optimization. The initial winter field tests were successfully conducted with embedded magnetic markers along the guardrails installed on the shoulders of Interstate-80 in the Sierra Mountain region close to Donner Summit, CA, during the winter of 2005.

The vehicle collision warning system (CWS) is an important research and application subject for vehicle safety. Most of this topic's research focuses on autonomous CWSs, where each vehicle detects potential collisions based entirely on the information measured by itself. Recently, an alternative scenario has arisen. This scenario is known as cooperative driving, where either the vehicle or the infrastructure can communicate its location, intention, or other information to surrounding vehicles or nearby infrastructure. Since installing a low-cost global-positioning-system (GPS) unit is becoming a common practice in vehicle applications, its implications in cooperative driving and vehicle safety deserve closer investigation. Furthermore, the future trajectory prediction may lead to a straightforward approach to detect potential collisions, yet its effectiveness has not been studied. This paper explores the engineering feasibility of a future-trajectory-prediction-based cooperative CWS when vehicles are equipped with a relatively simple differential GPS unit and relatively basic motion sensors. The goals of this paper are twofold: providing an engineering argument of possible functional architectures of such systems and presenting a plausible example of the proposed future-trajectory-based design, which estimates and communicates vehicle positions and predicts and processes future trajectories for collision decision making. In this paper, common GPS problems such as blockage and multipath, as well as common communication problems such as dropout and delays, are assumed. However, specific choices of GPS devices and communication protocol or systems are not the focus of this paper

This paper addresses the feasibility of post-impact maneuvers that aim to mitigate accident sequences. In particular, the discussions emphasize the possibility of applying steering control to stabilize the trajectories of vehicles involved in a collision. The problem of interest is defined with descriptions of accident scenarios, limitations of applicability, and performance requirements. A model of vehicle dynamics is constructed to facilitate the design of steering controllers. Various collision scenarios are simulated to demonstrate the effectiveness of a generic feedback controller. The analysis provided herein shows steering control is promising as a driver-assistance function in emergency situations.