| 1. |
- Andreotti, Eleonora, 1988, et al.
(författare)
-
Mathematical Definitions of Scene and Scenario for Analysis of Automated Driving Systems in Mixed-Traffic Simulations
- 2021
-
Ingår i: IEEE Transactions on Intelligent Vehicles. - 2379-8858. ; 6:2, s. 366-375
-
Tidskriftsartikel (refereegranskat)abstract
- This paper introduces a unified mathematical definition for describing commonly used terms encountered in systematical analysis of automated driving systems in mixed-traffic simulations. The most significant contribution of this work is in translating the terms that are clarified previously in literature into a mathematical set and function based format. Our work can be seen as an incremental step towards further formalisation of Domain-Specific-Language (DSL) for scenario representation. We also extended the previous work in the literature to allow more complex scenarios by expanding the model-incompliant information using set-theory to represent the perception capacity of the road-user agents. With this dynamic perception definition, we also support interactive scenarios and are not limited to reactive and pre-defined agent behavior. Our main focus is to give a framework to represent realistic road-user behavior to be used in simulation or computational tool to examine interaction patterns in mixed-traffic conditions. We believe that, by formalising the verbose definitions and extending the previous work in DSL, we can support automatic scenario generation and dynamic/evolving agent behavior models for simulating mixed traffic situations and scenarios. In addition, we can obtain scenarios that are realistic but also can represent rare-conditions that are difficult to extract from field-tests and real driving data repositories.
|
|
| 2. |
- Andreotti, Eleonora, 1988, et al.
(författare)
-
Potential impact of autonomous vehicles in mixed traffic from simulation using real traffic flow
- 2023
-
Ingår i: Journal of Intelligent and Connected Vehicles. - 2399-9802. ; 6:1, s. 1-15
-
Tidskriftsartikel (refereegranskat)abstract
- This work focuses on the potential impacts of the autonomous vehicles in a mixed traffic condition represented in traffic simulator Simulation of Urban MObility (SUMO) with real traffic flow. Specifically, real traffic flow and speed data collected in 2002 and 2019 in Gothenburg were used to simulate daily flow variation in SUMO. In order to predict the most likely drawbacks during the transition from a traffic consisting only manually driven vehicles to a traffic consisting only fully-autonomous vehicles, this study focuses on mixed traffic with different percentages of autonomous and manually driven vehicles. To realize this aim, several parameters of the car following and lane change models of autonomous vehicles are investigated in this paper. Along with the fundamental diagram, the number of lane changes and the number of conflicts are analyzed and studied as measures for improving road safety and efficiency. The study highlights that the autonomous vehicles' features that improve safety and efficiency in 100% autonomous and mixed traffic are different, and the ability of autonomous vehicles to switch between mixed and autonomous driving styles, and vice versa depending on the scenario, is necessary.
|
|
| 3. |
- Andreotti, Eleonora, 1988, et al.
(författare)
-
Safety-centred analysis of transition stages to traffic with fully autonomous vehicles
- 2020
-
Ingår i: IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC.
-
Konferensbidrag (refereegranskat)abstract
- The aim of this paper is to highlight and investigate the effects of increasing presence rate of autonomous vehicles (AVs) in terms of traffic safety and traffic flow characteristics. For this purpose, using existing driver models in traffic simulator software SUMO we identify and analyze those parameters that characterize and distinguish AVs' driving from manual driving in a heterogeneous traffic context. While it is essential to identify the parameters for traffic flow characteristics of heterogeneous fleets compared to homogeneous ones comprising manually driven vehicles (MV) only (i.e. current status), the safety aspects must be also accounted for. In order to combine these two fundamental aspects of heterogeneous traffic, we used a complete description of a highway driving scenario. The scenario integrates the perceptions of different type of vehicles (i.e. AV and MV) involved and the reaction times of human drivers and decision-making units of autonomous vehicles, to explore the impact of both the rate of AV presence and the perturbation in perception capabilities in highway scenarios.
|
|
| 4. |
- Andreotti, Eleonora, 1988, et al.
(författare)
-
Simulation-based impact projection of autonomous vehicle deployment using real traffic flow
- 2021
-
Tidskriftsartikel (övrigt vetenskapligt/konstnärligt)abstract
- In this work we focus on future projected impacts of the autonomous vehicles in a realistic condition representing mixed traffic. By using real flow and speed data collected in 2002 and 2019 in the city of Gothenburg, we replicated and simulated the daily flow variation in SUMO. The expansion of the city in recent years was reflected in an increase in road users, and it is reasonable to expect it will increase further. Through simulations, it was possible to project this increase and to predict how this will impact the traffic in future. Furthermore, the composition of vehicle types in the future traffic can be expected to change through the introduction of autonomous vehicles. In order to predict the most likely drawbacks during the transition from a traffic consisting only manually driven vehicles to a traffic consisting only fully-autonomous vehicles, we focus on mixed traffic with different percentages of autonomous and manually driven vehicles. To realize this aim, several parameters of the car following and lane change models of autonomous vehicles are investigated in this paper. Along with the fundamental diagram, the number of lane changes and the number of conflicts are analyzed and studied as measures for improving road safety and efficiency.
|
|
| 5. |
- Bayraktar, Ertugrul, et al.
(författare)
-
A hybrid image dataset toward bridging the gap between real and simulation environments for robotics
- 2019
-
Ingår i: Machine Vision and Applications. - : Springer Science and Business Media LLC. - 1432-1769 .- 0932-8092. ; 30:1, s. 23-40
-
Tidskriftsartikel (refereegranskat)abstract
- The primary motivation of computer vision in the robotics field is to obtain a perception level that is as close as possible to human visual system. To achieve this, the inclusion of large datasets is necessary, sometimes involving less-frequent and seemingly irrelevant data to increase the system robustness. To minimize the effort and time in forming such extensive datasets from real world, the preferred method is to utilize simulation environments, replicating real-world conditions as much as possible. Following this solution path, the machine vision problems in robotics (i.e., object detection, recognition, and manipulation) often employ synthetic images in datasets and, however, do not mix them with real-world images. When the systems are trained only using the synthetic images and tested within the simulated world, the tasks requiring object recognition in robotics can be accomplished. However, the systems trained using this procedure cannot be directly used in the real-world experiments or end-user products due to the inconsistencies between real and simulation environments. Therefore, we propose a hybrid image dataset including annotated desktop objects from real and synthetic worlds (ADORESet). This hybrid dataset provides purposeful object categories with a sufficient number of real and synthetic images. ADORESet is composed of colored images with the dimension of 300×300 pixels within 30 categories. Each class has 2500 real-world images acquired from the wild web and 750 synthetic images that are generated within Gazebo simulation environment. This hybrid dataset enables researchers to implement their own algorithms for both real-world and simulation environment conditions. ADORESet is composed of fully annotated object images. The limits of objects are manually specified, and the bounding box coordinates are provided. The successor objects are also labeled to give statistical information and the likelihood about the relations of the objects within the dataset. To further demonstrate the benefits of this dataset, it is tested in object recognition tasks by fine-tuning the state-of-the-art deep convolutional neural networks such as VGGNet, InceptionV3, ResNet, and Xception. The possible combinations regarding the data types for these models are compared in terms of time, accuracy, and loss values. As a result of the conducted object recognition experiments, training with all-real images yields approximately 49% validation accuracy for simulation images. When the training is performed with all-synthetic images and validated using all-real images, the accuracy becomes lower than 10%. If the complete ADORESet is employed for training and validation, the hybrid dataset validation accuracy reaches approximately to 95%. This result proves further that including the real and synthetic images together in the training and validation sessions increases the overall system accuracy and reliability.
|
|
| 6. |
- Bayraktar, Ertugrul, et al.
(författare)
-
Analysis of feature detector and descriptor combinations with a localization experiment for various performance metrics
- 2017
-
Ingår i: Turkish Journal of Electrical Engineering and Computer Sciences. - : The Scientific and Technological Research Council of Turkey (TUBITAK-ULAKBIM) - DIGITAL COMMONS JOURNALS. - 1300-0632 .- 1303-6203. ; 25
-
Tidskriftsartikel (refereegranskat)abstract
- The purpose of this study is to give a detailed performance comparison about the feature detector and descriptor methods, particularly when their various combinations are used for image matching. As the case study, the localization experiments of a mobile robot in an indoor environment are given. In these experiments, 3090 query images and 127 dataset images are used. This study includes five methods for feature detectors such as features from accelerated segment test (FAST), oriented FAST and rotated binary robust independent elementary features (BRIEF) (ORB), speeded-up robust features (SURF), scale invariant feature transform (SIFT), binary robust invariant scalable keypoints (BRISK), and five other methods for feature descriptors which are BRIEF, BRISK, SIFT, SURF, and ORB. These methods are used in 23 different combinations and it was possible to obtain meaningful and consistent comparison results using some performance criteria defined in this study. All of these methods are used independently and separately from each other as being feature detector or descriptor. The performance analysis shows the discriminative power of various combinations of detector and descriptor methods. The analysis is completed using five parameters such as (i) accuracy, (ii) time, (iii) angle difference between keypoints, (iv) number of correct matches, and (v) distance between correctly matched keypoints. In a range of 60°, covering five rotational pose points for our system, FAST-SURF combination gave the best results with the lowest distance and angle difference values and highest number of matched keypoints. The combination SIFT-SURF is obtained as the most accurate combination with 98.41% of correct classification rate. The fastest algorithm is achieved with ?ORB-BRIEF? combination with a total running time 21303.30 seconds in order to match 560 images captured during the motion with 127 dataset images.
|
|
| 7. |
- Bayraktar, Ertugrul, et al.
(författare)
-
Object manipulation with a variable-stiffness robotic mechanism using deep neural networks for visual semantics and load estimation
- 2020
-
Ingår i: Neural Computing and Applications. - : Springer Science and Business Media LLC. - 0941-0643 .- 1433-3058. ; 32:13, s. 9029-9045
-
Tidskriftsartikel (refereegranskat)abstract
- In recent years, the computer vision applications in the robotics have been improved to approach human-like visual perception and scene/context understanding. Following this aspiration, in this study, we explored the possibility of better object manipulation performance by connecting the visual recognition of objects to their physical attributes, such as weight and center of gravity (CoG). To develop and test this idea, an object manipulation platform is built comprising a robotic arm, a depth camera fixed at the top center of the workspace, embedded encoders in the robotic arm mechanism, and microcontrollers for position and force control. Since both the visual recognition and force estimation algorithms use deep learning principles, the test set-up was named as Deep-Table. The objects in the manipulation tests are selected from everyday life and are common to be seen on modern office desktops. The visual object localization and recognition processes are performed from two distinct branches by deep convolutional neural network architectures. We present five of the possible cases, having different levels of information availability on the object weight and CoG in the experiments. The results confirm that using our algorithm, the robotic arm can move different types of objects successfully varying from several grams (empty bottle) to around 250 g (ceramic cup) without failure or tipping. The proposed method also shows that connecting the object recognition with load estimation and contact point further improves the performance characterized by a smoother motion.
|
|
| 8. |
- Boyraz Baykas, Pinar, 1981
(författare)
-
Acoustic road-type estimation for intelligent vehicle safety applications
- 2014
-
Ingår i: International Journal of Vehicle Safety. - 1479-3105 .- 1479-3113. ; 7:2, s. 209-222
-
Tidskriftsartikel (refereegranskat)abstract
- A low-cost acoustic road-type classification system is proposed to be used in road-tyre friction force estimation in active safety applications. The system employs audio signal processing and extracts features such as linear predictive coefficients (LPC), mel-frequency cepstrum coefficients (MFCC) and power spectrum coefficients (PSC). The features are extracted using time windows of 0.02, 0.05 and 0.1 seconds in order to find the best representative window for the signal properties which should also be as short as possible for active safety systems. In order to find the best feature space, a variance analysis based approach is considered to represent the road types as distinguished classes. Optimised feature space is classified using artificial neural networks (ANN). The results show that the designed ANN can classify the road types with 91% accuracy at worst condition. To demonstrate the value of the system, a case study including traction control application is reported.
|
|
| 9. |
- Boyraz Baykas, Pinar, 1981, et al.
(författare)
-
Active Accident Avoidance Case Study: Integrating Drowsiness Monitoring System with Lateral Control and Speed Regulation in Passenger Vehicles
- 2008
-
Ingår i: IEEE International Conference on Vehicular Electronics and Safety. - 9781424423590 ; , s. 293-298
-
Konferensbidrag (refereegranskat)abstract
- This study proposes architecture for integrating intelligent control systems into vehicles, with special consideration to include the human-driver in the control loop. As a case study of the proposed architecture, drowsiness monitoring system is combined with an adaptive and robust lateral controller. Drowsiness is considered to be related to the uncertainty in steering wheel commands for the vehicle lateral movement. Using a robust control theory scheme, the uncertainties from road-vehicle forces and driver inputs are addressed resulting in a lateral controller. The controller is able to re-shape the frequency response of the vehicle in both lateral acceleration and side-slip angle, shifting the response into more stable areas in Nyquist diagram. An additional speed reduction finalizes the complete stabilization of the vehicle-driver system. The stabilization in lateral dynamics of the car and speed reduction addresses the characteristics of the road accident patterns including drowsy/sleepy drivers.
|
|
| 10. |
- Boyraz Baykas, Pinar, 1981, et al.
(författare)
-
Active vehicle safety system design based on driver characteristics and behaviour
- 2009
-
Ingår i: International Journal of Vehicle Safety. - 1479-3105 .- 1479-3113. ; 4:4, s. 330-364
-
Tidskriftsartikel (refereegranskat)abstract
- In the development of driver adaptive and context aware active safety applications, driver-vehicle interaction signals offer excellent opportunities for advanced system design, yet limited progress has been realised. The implementation of driver adaptive and context aware systems requires longer time windows to analyse the current status of the driver and/or traffic situation ahead. In this study, a summary of systems that can be realised based on the long-term analysis of driver-vehicle interaction signals is presented. These signals are readily obtained by using Controller Area Network (CAN) Bus via On Board Diagnostic System (OBD) II port that can be utilised at low cost. Based on the analysis results, quantitative metrics are suggested that can be used in many ways, with two prospects considered here: (1) manoeuvres can be recognised for context aware intelligent active safety and (2) the models or signal processing methods can be proposed so as to distinguish distracted/impaired driver behaviour from normal/safe behaviour.
|
|
