Prototype Autonomous Vehicle Research Articles

A barrier to innovation: Europe’s ad-hoc cross-border framework for testing prototype autonomous vehicles

ABSTRACTThe conglomeration of regulatory frameworks for the testing of prototype autonomous vehicles in Europe creates a challenging task for developers and researchers planning pilots across borders. While there are examples of international autonomous driving projects and cooperation in autonomous vehicle research, Europe lacks a mutually recognised testing procedure for autonomous vehicle pilots, and incompatible legal and administrative processes in each country creates a disincentive for ambitious cross-border testing. The diverse climate and topography of Europe potentially provides a rigorous testing ground for autonomous vehicles, and an opportunity to prepare the new technology to deal with varied signage, language and driver behaviour encountered when travelling across multiple countries. Prototype vehicles tested in such conditions provide valuable insight for research and product development. This may be encouraged by a more harmonised prototype testing framework including a pan-European type-approval exemption scheme for prototype vehicles, and for cross-border tests to be coordinated by regional organisations interested in promoting development in border areas.

Augmented Vehicular Reality

Autonomous vehicle prototypes today come with line-of-sight depth perception sensors like 3D cameras. These 3D sensors are used for improving vehicular safety in autonomous driving, but have fundamentally limited visibility due to occlusions, sensing range, and extreme weather and lighting conditions. To improve visibility and performance, we explore a capability called Augmented Vehicular Reality (AVR). AVR broadens the vehicle's visual horizon by enabling it to wirelessly share visual information with other nearby vehicles. We show that AVR is feasible using off-the-shelf wireless technologies, and it can qualitatively change the decisions made by autonomous vehicle path planning algorithms. Our AVR prototype achieves positioning accuracies that are within a few percentages of car lengths and lane widths, and it is optimized to process frames at 30fps.

Reactive navigation of autonomous vehicle

Research on dynamic path planning for autonomous vehicle has been increased significantly in recent years because of the percentage of car accidents reduced by autonomous driving. The functions of Artificial Potential Fields (APF) as mentioned in previous studies, particularly nonlinear constraint, speed and safe distance, are seldom discussed or mentioned. In this paper, we focus on the need for vehicle safety. Thus, this study proposes a fuzzy artificial potential field path planning algorithm (FZ-APF) for local path planning in vehicle. FZ-APF introduces the distance factor to solve the well-known drawbacks of the traditional APF like goal non-reachable with obstacle nearby (GNRON) problem. However, FZ-APF not only optimizes path but also deals with the constraints of vehicle like the limit of steering angle at different speeds. Therefore, FZ-APF introduces longitudinal potential function to make the planned path accord with vehicle characteristics. In addition, to achieve more safe and efficient FZ-APF for vehicle, the fuzzy technique which adjusts the parameters at any moment according to the current speed and safety distance of the vehicle is added into FZ-APF. Numerical simulations are presented and validate the proposed algorithm under some complicated test scenarios. Experiment eventually illustrates the efficiency of this FZ-APF algorithm using an autonomous vehicle prototype.

Improving steering convergence in autonomous vehicle steering control

<p>Steering control is a critical design element in autonomous vehicle development since it will determine whether the vehicle can navigate safely or not. For the prototype of UiTM Autonomous Vehicle 0 (UiTM AV0), Vexta motor is used to control the steering whereas Pulse Width Modulation (PWM) signal is responsible to drive the motor. However, by using PWM signal it is difficult to converge to the desired steering angle and furthermore time taken for steering angle to converge is much longer. Thus, Proportional Integral Derivative (PID) has been introduced in this autonomous vehicle steering controller to improve the convergence of the steering. Meanwhile a microcontroller was used to control the Vexta Motor direction and perform the calculation of the desired steering angle. Simulation results showed PID controller showed better time taken and preicison of successful convergence of the desired steering angle compared to the PWM controller. Analysis results showed that PID controller significantly reduce the overshooting of steering angle and significantly improve the time taken for convergence by up to 37 seconds faster than PWM controller in UiTM AV0.</p>

Enhanced emotion enabled cognitive agent-based rear-end collision avoidance controller for autonomous vehicles

Amongst collisions, rear-end collisions are the deadliest. Several rear-end collision avoidance solutions have been proposed recently in the literature. A key problem with existing solutions is their dependence on precise mathematical models. However, real world driving is influenced by a number of nonlinear factors. These include road surface conditions, driver reaction time, pedestrian flow, and vehicle dynamics. These factors involve so many different variations that precise mathematical solutions are hard to obtain, if not impossible. This problem with precise control-based rear-end collision avoidance schemes has also previously been addressed using fuzzy logic, but the excessive number of fuzzy rules straightforwardly prejudices their efficiency. Furthermore, such fuzzy logic-based controllers have been proposed without the use of an appropriate modeling technique. One such modeling technique is agent-based modeling. This technique is suitable because it allows for mimicking the functions of an artificial human driver executing fuzzy rules. Keeping in view these limitations, we propose an enhanced emotion enabled cognitive agent (EEEC_Agent)-based controller. The proposed EEEC_Agent helps autonomous vehicles (AVs) avoid rear-end collisions with fewer rules. One key innovation in its design is to use the human emotion of fear. The resultant agent is very efficient and also uses the Ortony–Clore–Collins (OCC) model. The fear generation mechanism of EEEC_Agent is verified through NetLogo simulation. Furthermore, practical validation of EEEC_Agent functions is performed by using a specially built prototype AV platform. Finally, a qualitative comparison with existing state-of-the-art research works reflects that the proposed model outperforms recent research proposals.

A Bio-Inspired Cognitive Agent for Autonomous Urban Vehicles Routing Optimization

Autonomous urban vehicle prototypes are expected to be efficient even in not explicitly planned circumstances and dynamic environments. The development of autonomous vehicles for urban driving needs real-time information from vehicles and road network to optimize traffic flows. In traffic agent-based models, each vehicle is an agent, while the road network is the environment. Cognitive agents are able to reason on the perceived data, to evaluate the information obtained by reasoning, and to learn and respond, preserving their self-sufficiency, independency, self-determination, and self-reliance. In this paper, a bio-inspired cognitive agent for autonomous urban vehicles routing optimization is proposed. The use of selected bio-inspired analyzing techniques, which are commonly employed to investigate the topological and functional features of a metabolic network, allows the agent to analyze the structural aspects of a road network, find its extreme pathways and outline the balanced flow combinations. This approach optimizes traffic flows over network, minimizes road congestions, and maximizes the number of autonomous vehicles reaching their destination target. Agent behavior has been tested using data coming from Palermo urban road network, Italy, while the adopted bio-inspired analysis techniques have been compared with the A* literature algorithm. Experimental results demonstrate that the approach permits to find a better global routing optimization solution. To the best of our knowledge, it is the first time that metabolic mechanisms involved in a cell survival process have been used to design a congestion solution.

A functional reference architecture for autonomous driving

ContextAs autonomous driving technology matures toward series production, it is necessary to take a deeper look at various aspects of electrical/electronic (E/E) architectures for autonomous driving.ObjectiveThis paper describes a functional reference architecture for autonomous driving, along with various considerations that influence such an architecture. The functionality is described at the logical level, without dependence on specific implementation technologies.MethodEngineering design has been used as the research method, which focuses on creating solutions intended for practical application. The architecture has been refined and applied over a 5 year period to the construction of prototype autonomous vehicles in three different categories, with both academic and industrial stakeholders.ResultsThe architectural components are divided into categories pertaining to (i) perception, (ii) decision and control, and (iii) vehicle platform manipulation. The architecture itself is divided into two layers comprising the vehicle platform and a cognitive driving intelligence. The distribution of components among the architectural layers considers two extremes: one where the vehicle platform is as “dumb” as possible, and the other, where the vehicle platform can be treated as an autonomous system with limited intelligence. We recommend a clean split between the driving intelligence and the vehicle platform. The architecture description includes identification of stakeholder concerns, which are grouped under the business and engineering categories. A comparison with similar architectures is also made, wherein we claim that the presence of explicit components for world modeling, semantic understanding, and vehicle platform abstraction seem unique to our architecture.ConclusionThe concluding discussion examines the influences of implementation technologies on functional architectures and how an architecture is affected when a human driver is replaced by a computer. The discussion also proposes that reduction and acceleration of testing, verification, and validation processes is the key to incorporating continuous deployment processes.

Safe and reliable navigation in crowded unstructured pedestrian areas

In this paper, the navigation system of the autonomous vehicle prototype Verdino is introduced. Two navigation levels are considered. In the first level, a trajectory is generated from the current position toward a goal that considers two different approaches. In the first, the minimum cost path is obtained using a classical approach (used for regular navigation). The second approach is a little more complex, relying on a set of precomputed primitives representing the motion model of the vehicle, which are used as part of an ARA⁎ algorithm in order to find the best trajectory. This trajectory consists of both forward and backward motion segments for complex maneuvers. In the second level, a local planner is in charge of computing the commands sent to the vehicle in order to follow the trajectory. A set of tentative local trajectories is computed in the Frenét space and scored using several factors, described in this paper. Some results for the two navigation levels are shown at the end of this document. For the global planner, several examples of the maneuvers obtained are shown and certain related factors are quantified and compared. As for the local planner, a study on the influence of the defined weights on the vehicle׳s final behavior is presented. Also, from these tests several configurations have been chosen and ranked according to two different proposed behaviors. The navigation system shown has been tested both in simulated and in real conditions, and the attached video shows the vehicle׳s real-world performance.

Extensive Tests of Autonomous Driving Technologies

This paper presents the vision of the Artificial Vision and Intelligent Systems Laboratory (VisLab) on future automated vehicles, ranging from sensor selection up to their extensive testing. VisLab's design choices are explained using the BRAiVE autonomous vehicle prototype as an example. BRAiVE, which is specifically designed to develop, test, and demonstrate advanced safety applications with different automation levels, features a high integration level and a low-cost sensor suite, which are mainly based on vision, as opposed to many other autonomous vehicle implementations based on expensive and invasive sensors. The importance of performing extensive tests to validate the design choices is considered to be a hard requirement, and different tests have been organized, including an intercontinental trip from Italy to China. This paper also presents the test, the main challenges, and the vehicles that have been specifically developed for this test, which was performed by four autonomous vehicles based on BRAiVE's architecture. This paper also includes final remarks on VisLab's perspective on future vehicles' sensor suite.

What Will the Law Do About Autonomous Vehicles?

Autonomous vehicles are just beginning to emerge on roads and highways all over the world. The autonomous vehicle prototypes available now provide some clues to understanding how law will apply to this developing technology. There appear to be several types of autonomous vehicles. Most of the experimental autonomous vehicles tend to be either self-contained or interconnected or some combination of these two types. Another category of autonomous vehicles, a self-willed autonomous vehicle, appears only in fiction - often fantasy or horror stories about vehicles run amok or falling in love with humans. More benign self-willed vehicles include Herbie the Love Bug and Chitty Chitty Bang Bang. Until the singularity, when machines are smarter than humans are, self-willed autonomous vehicles are likely to exist in our imaginations, rather than on the roads.It is not clear what the market for autonomous vehicles will be. Driving enthusiasts will probably avoid them. However, differently abled persons, such as persons with disabilities, as well as children and the elderly are likely to find autonomous vehicles especially attractive.Autonomous vehicles may possibly appeal to commuters, who find long daily commutes over boring roads tedious. However, whether ordinary consumers will be inclined to purchase autonomous vehicles remains a matter of speculation.In very general terms, autonomous vehicles are self-driving cars operated by artificial intelligence using automated controls and, usually, GPS. Internal vehicle sensors constantly monitor the car’s mechanical operation - such as its speed, steering, braking, and the like. The differences between the self-contained and interconnected autonomous vehicles lie in the ways in which the two types assess the dynamic driving environment on the roads where they travel.Self-contained autonomous vehicles gather and keep all information within the vehicle. They collect data from outward facing sensors (cameras, lidar, radar, etc.) to visualize and predict the changing roadway situation around the moving car. Most of the self-contained autonomous vehicle prototypes require that roadways on which these vehicles travel be mapped in advance by a human driver.On the other hand, the interconnected type of autonomous vehicle relies on real-time communications from other vehicles or the roadside to provide data about what is happening in the roadway environment in which it operates. Interconnected autonomous vehicles do not require pre-mapped routes, but rather rely on data about the driving environment that is communicated into the vehicle from outside sources.These technical differences result in a number of interesting legal questions that may have different answers depending on the type of autonomous vehicle. How should legal rules determine civil and criminal legal liability, for example, when an autonomous vehicle breaks a traffic rule or damages another’s vehicle or is involved in vehicular manslaughter? Products liability could be placed on the manufacturer or seller of the vehicle, on the hardware provider, or on the maker of the software used by the vehicle’s artificial intelligence. Privacy rights, insurance law, and regulatory jurisdiction have yet to be determined. As autonomous vehicles’ new, innovative transportation technology develops, the law also will need to innovate.

Diseño de AUV.Arquitectura de hardware y software

This paper discusses the strategy under which were conceived the hardware and software architecture for autonomous vehicle prototype: HRC-AUV, and the selection of the fundamental elements that compose them. The obtained design weights simplicity and development in terms of low cost, factors useful to researchers begin their activity in this field. The paper summarizes the benefits provided by these structures and preliminary operational tests that have been submitted to demonstrate the validity of their use in the operation of an AUV. Likewise are linearized dynamic models of the plant, used in the tuning of the control loops are presented. The response of such loops and in general the HRC-AUV navigating in the ocean is presented through the results of several experimental tests.

A novel fused neural network controller for lateral control of autonomous vehicles

We present a novel fused feed-forward neural network controller inspired by the notion of task decomposition principle. The controller is structurally simple and can be applied to a class of control systems that their control requires manipulation of two input variables. The benchmark problem of inverted pendulum is such example that its control requires availability of the angle as well as the displacement. We demonstrate that the lateral control of autonomous vehicles belongs to this class of systems and successfully apply the proposed controller to this problem. The parameters of the controller are encoded into real value chromosomes for genetic algorithm (GA) optimization. The neural network controller contains three neurons and six connection weights implying a small search space implying faster optimization time due to few controller parameters. The controller is also tested on two benchmark control problems of inverted pendulum and the ball-and-beam system. In particular, we apply the controller to lateral control of a prototype semi-autonomous vehicle. Simulation results suggest a good performance for all the tested systems. To demonstrate the robustness of the controller, we conduct Monte-Carlo evaluations when the system is subjected to random parameter uncertainty. Finally experimental studies on the lateral control of a prototype autonomous vehicle with different speed of operation are included. The simulation and experimental studies suggest the feasibility of this controller for numerous applications.

Modeling and Simulation of an Autonomous Hybrid- Electric Military Vehicle

The U.S. Army TACOM-TARDEC developed and validated a high-fidelity six-degree-of-freedom model to use in a trade study for the development of a prototype autonomous vehicle. The model captures realistic dynamics of the six-wheeled, skid-steered vehicle along with the electrical, thermal, and mechanical response of a detailed series hybrid-electric power system with in-hub drive motors, lithium-ion battery, and generator linked to a diesel engine. These components were modeled and integrated via extensive power and energy component libraries developed for use with a high-fidelity software tool for dynamics modeling. Further, the vehicle model’s entire complement of components was integrated in a flexible configuration that allowed them to be readily adjusted or swapped out so the user could use the model to ascertain the relative effects of modifying the vehicle’s structural or power system components on specific vehicle evaluation criteria. Such criteria include the vehicle’s performance with high-speed stability, skid steering stability, body pitch/roll/dive/squat characteristics, braking capability, road/soft-soil traversal, and steering maneuverability.The model captures both the on- and off-road mobility for the vehicle via use of an extensive library of various terrains including hard surface, sand, sandy loam, clay soil, and snow. Further, detailed waypointbased path navigation routines automate the vehicle’s traversal over a number of user-selectable courses including some established military courses such as Churchville-B, Perryman 1, 3, and A, and Munson with user-defined vehicle velocities. The model functions as an executable file run independent of any proprietary or close-source software; the user utilizes a simplified interface to vary any of the variables associated with the vehicle’s geometry, power system, course and speed to navigate, and terrain type applied to the course. The graphical view for the vehicle traversing the selected terrain is shown with an open source 3D graphics tool. The model was validated by applying the specifications in the model for the prototype vehicle of the first-generation of autonomous six-wheeled skid-steered vehicle, simulating the model in maneuvers identical to those the prototype vehicle performed, and comparing the simulated and actual results; the data matched and the model was successfully validated.The vehicle model was designed primarily for the trade study for the design of a specific vehicle, but was created with sufficient flexibility and capability for modeling future vehicles as well. The interchangeability of the vehicle models’ components and environments allow a user to modify or replace the vehicle’s power system components, chassis masses, tires, transmission, duty cycles, courses to traverse, and many other aspects of the vehicle. Thus the user can essentially model any vehicle with similar types of components or structures and use that model to determine the impact of those elements upon many vehicle design considerations such as mass requirements, volume constraints, power system requirements, wheels design, suspension characteristics, and controls. Several new vehicle models are already being developed using this model’s flexibility and capability.

Springrobot: A Prototype Autonomous Vehicle and Its Algorithms for Lane Detection

This work presents the current status of the Springrobot autonomous vehicle project, whose main objective is to develop a safety-warning and driver-assistance system and an automatic pilot for rural and urban traffic environments. This system uses a high precise digital map and a combination of various sensors. The architecture and strategy for the system are briefly described and the details of lane-marking detection algorithms are presented. The R and G channels of the color image are used to form graylevel images. The size of the resulting gray image is reduced and the Sobel operator with a very low threshold is used to get a grayscale edge image. In the adaptive randomized Hough transform, pixels of the gray-edge image are sampled randomly according to their weights corresponding to their gradient magnitudes. The three-dimensional (3-D) parametric space of the curve is reduced to the two-dimensional (2-D) and the one-dimensional (1-D) space. The paired parameters in two dimensions are estimated by gradient directions and the last parameter in one dimension is used to verify the estimated parameters by histogram. The parameters are determined coarsely and quantization accuracy is increased relatively by a multiresolution strategy. Experimental results in different road scene and a comparison with other methods have proven the validity of the proposed method.

Underwater vehicle design using integrated plant-controller optimization methods

This paper addresses the problem of integrated design of autonomous underwater vehicle (AUV) plant parameters and feedback controllers to meet mission performance requirements with minimum energy expenditure. This research topichas been motivated by the fact that significant energy savings and increased dynamic performance can in principle be obtained if the process of control system design is integrated with the design of the vehicle itself, thus departing considerably from the classical approach whereby the plant structure is essentially fixed a priori. As a first step towards the solution of this general problem, the following simpler problem is addressed and solved: given an AUV (with a fixed baseline configuration) that is required to operate over a finite set of representative trim conditions in the vertical plane in the course of a given mission scenario, determine the optimal size of the bow and stern planes so that the average propulsion power required to execute that mission is minimized. The minimization process is subjectto open-loop stability requirements, and to theexistence of stabilizing feedback controllers that can meet time and frequency closed-loop requirements about each trim condition. The paper introduces a methodology to solve this combined plant controller optimization problem in the framework of convex optimization theory and describes its application to theselection of the optimal size of bow and stern planes for a prototype autonomous vehicle.

Optimal AUV Control Surface Sizing Using Convex Optimization Methods ⋆

This paper addresses the problem of integrated design of AUV plant parameters and feedback controllers to meet mission performance requirements with minimum energy expenditure. Given an AUV that is required to operate over a finite set of representative trim conditions in the vertical plane in the course of a given mission scenario, the objective is to determine the optimal size of the bow and stern planes so that the average propulsion power required to execute that mission is minimized. The minimization process is subject to open loop stability requirements, and to the existence of stabilizing feedback controllers that can meet time and frequency closed loop requirements about each trim condition. The paper introduces a methodology to solve this combined plant/controller optimization problem in the framework of convex optimization theory, and describes its application to the selection of the optimal size of bow and stern planes for a prototype autonomous vehicle.

海洋観測用シャトル型海中ロボットＡＬＢＡＣ

The ALBAC is a prototype of autonomous underwater vehicle which was designed and constructed for oceanographic measurement of water column on the way and from seabed. The ALBAC has a pair of wings which provide large lift force to move horizontally without consuming energy of batteries. Major equipment including control actuators is installed in one pressure hull. Since these apparatuses are not exposed to the sea water, the actuators are not effected by environment resulting in significantly high reliability.To ensure the total system performance, experiments in sea were conducted and distribution of water temperature was simultaneously measured.

Knowledge-Based Approach Toward Developing An Autonomous Helicopter System

The advent of advanced computer architectures for parallel and symbolic processing has evolved technology to the point at which prototype autonomous vehicles are being developed. Control of such devices requires communication between knowledge-based subsystems in charge of the vision, planning, and control aspects necessary to make autonomous systems func-tional in a real-world environment. The performance of autonomous vehicle systems is currently limited by their inability to accurately analyze their surrounding environment. In order to function in dynamic situations, autonomous vehicles must be capable of interpreting terrain on the basis of predetermined mission goals. This paper describes an autonomous airborne-vehicle simulation currently being developed at the Georgia Tech Research Institute. The Autonomous Helicopter System (AHS) is a multimission system consisting of three distinct sections: vision, planning, and control. The vision section provides the local and global scene analyses that are symbolically represented and passed to the planning section as the initial route-planning constraints. The planning section generates a task-dependent path for the vehicle to traverse that assures maximum mission system successes as well as survivability. The control section validates the path and either executes the given route or feeds back to previous sections in order to resolve conflicts.

NEW DESIGNS OF UNDERWATER VEHICLES FROM UNDERWATER TECHNOLOGY DEPARTMENT, TECHNICAL UNIVERSITY OF GDA SK

New designs have been recently added to wide range of submersibles developed at Underwater Technology Department. After completion of KORAL AT ROY, an autonomous vehicle prototype, for Baltic Sea environment monitoring, is being currently developed. Paper describes operational requirements for the craft and its principal components Vehicle control by remote means is not discarded however This technology has been adopted for a sea mine destruction system with disposable low cost vehicle used for mine identification and destruction. Other interesting development is the newest version of diver towing vehicle. The vehicle is constantly improved and modified to different requirements regarding range and special equipment. The latest design is shortly described.