Mini autonomous car architecture for urban driving scenarios

Karabulut, Gökhan
Autonomous cars capable of driving in city traffic have been long studied in architectures decomposed into perception, planning, and control components. Recent advances in deep learning techniques considerably contributed to the perception component of this approach. These techniques also laid the groundwork for the progress of other approaches such as end-to-end learning of steering commands and driving affordances from camera images. Though these approaches are promising to simplify the overall architecture, the decomposed architectures are found more persuasive, constituting the majority of today's state-of-the-art, market-oriented driverless cars. However, studies on small-scale autonomous cars, which are considered low-cost and rapid prototyping platforms, are not on a par with research on the modern decomposed architectures. These studies often remain limited to end-to-end approaches or resort to traditional image processing techniques in over-simplified traffic scenarios. In this thesis, we present a decomposed architecture for small-scale cars covering extended traffic scenarios with seven traffic signs, traffic lights, lane changes, cloverleaf interchange, pedestrian crossings, and parking. To realize this architecture, we created segmentation and classification datasets. We trained two deep learning models for learning lane semantics and classifying traffic signs and lights. We developed a behavior planner to decide on the best behavior primitives for traffic scenes. Based on these behavior primitives, we implemented a trajectory planner to find optimal trajectories along the lanes and a controller to follow these trajectories. With our novel lane segmentation scheme, 97% accurate classifier, robust planner and controller algorithms, we achieved successful drives on simulated and real courses.


ÖZTÜRK, İbrahim; Özkan, Türker; WALLÉN WARNER, Henriette; Department of Psychology (2021-7)
With technological developments, vehicles with different capabilities are becoming part of the traffic system. In recent years, vehicles with different levels of automation are taking the attention of both industry and academia. In addition, traffic climate, traffic locus of control and driving skills have been related to various road safety outcomes such as accidents. The present study examines how traffic climate, traffic locus of control and driving skills are related to drivers’ automated vehicle prefer...
Analysis of comprehension of traffic signs: a pilot study in Ankara, Turkey
Kırmızıoğlu, Erkut; Tüydeş Yaman, Hediye; Department of Civil Engineering (2010)
Traffic signs, which are extremely important for traffic safety, aims to regulate traffic by providing information about the characteristics of road and road environment for drivers. The success of traffic signs mainly rely on the easy comprehensibility of its meaning in a short time. Further more, today’s global economies and transportation systems emphasize the need for more universial traffic signs which was the main motivation of two main treaties on traffic signs; Vienna Convetion in 1968 and European ...
Trajectory planning and tracking for autonomous vehicles
Çiçek, Haluk Levent; Schmidt, Klaus Verner; Department of Electrical and Electronics Engineering (2022-12-27)
Finding appropriate paths is an essential issue for the development of autonomous vehicles and robots. Hereby, it has to be considered that autonomous vehicles cannot follow sharp corners, as they cannot turn on a single point. Therefore, it is important to compute smooth paths that have additional desirable properties such as minimum length and sufficient distance from obstacles. Furthermore, practical applications require the computation of such paths in real time. This thesis develops a general method...
Detection of urban traffic patterns from Floating Car Data (FCD)
ALTINTAŞI, ORUÇ; Tüydeş Yaman, Hediye; Tuncay, Kağan (Elsevier BV; 2016-09-07)
Real time data collection in traffic engineering is crucial for better traffic corridor control and management. In the literature, many data collection methods have been used such as; magnetic loops, road tube counters, piezo sensors, radars, Bluetooth etc. to estimate the link occupancy, average speed or density of a corridor. More recently, Floating Car Data (FCD) has become another important traffic data source and has an increasing usage due to its lower cost and higher coverage despite its reliability ...
Mobility-on-demand scenarios relying on lightweight autonomous and connected vehicles for large pedestrian areas and intermodal hubs
Kumru, Murat; Makarem, Laleh; Gillet, Dens (null; 2017-10-05)
This paper presents Mobility-On-Demand (MoD) scenarios relying on lightweight autonomous and connected vehicles to ease commuting in large pedestrian areas and intermodal hubs. In particular, we discuss the case of the EPFL campus as an example of a large pedestrian area and the case of the main Zurich train station as an example of a large intermodal hub. These two scenarios help to investigate the potential benefits and the implementation challenges of innovative lightweight autonomous and connected vehic...
Citation Formats
G. Karabulut, “Mini autonomous car architecture for urban driving scenarios,” Thesis (M.S.) -- Graduate School of Natural and Applied Sciences. Computer Engineering., Middle East Technical University, 2019.