Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Das Kapitel untersucht das transformative Potenzial der vollständigen selbstfahrenden Automatisierung in der urbanen Mobilität und hebt ihre Vorteile wie einen verringerten Energieverbrauch, geringeren Fahrstress, Kosteneinsparungen in der Logistik und verbesserte Verkehrssicherheit hervor. Er betont die bis Mitte des Jahrhunderts prognostizierte deutliche Verringerung der Verkehrstoten durch autonomes Fahren. Der Text untersucht die technologischen Herausforderungen in städtischen Umgebungen, einschließlich der Komplexität urbaner Umgebungen mit zahlreichen Akteuren und der Notwendigkeit ultrasicheren autonomen Fahrens der Stufen 3-4. Das Integra-Projekt, eine Zusammenarbeit zwischen spanischen F & E-Zentren, konzentriert sich auf die Entwicklung von Lösungen für autonomes Fahren in städtischen Umgebungen, wobei Bereiche wie Funktionen des autonomen Fahrens, Konnektivität, Insassensicherheit und autonome Warenlieferungen berücksichtigt werden. In diesem Kapitel wird die Bedeutung einer präzisen Wahrnehmung der Umwelt, die Rolle der V2X-Kommunikation bei der Verbesserung der Fahrsicherheit und die Notwendigkeit adaptiver Rückhaltesysteme zur Gewährleistung der Insassensicherheit diskutiert. Es umfasst auch die technische Charakterisierung städtischer Lieferwege und die Entwicklung intelligenter mobiler Schließfächer für den Güterverkehr. Die Schlussfolgerung unterstreicht die Schlüsselaspekte, die für den erfolgreichen Einsatz des autonomen Fahrens in städtischen Umgebungen erforderlich sind, darunter die präzise Wahrnehmung der Umwelt, effektive V2X-Kommunikation, angemessene Rückhaltesysteme und die Anpassung des Fahrzeugs an die Umgebungsbedingungen.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
It is an essential requisite in order to achieve safe autonomous driving in urban environments, to address the mobility from a multidisciplinary approach. In this sense, technological development is crucial in both passenger and goods transport. Within Integra project, in which the work described here is framed, autonomous and connected driving is addressed from a dual perspective: urban environments and logistics environments. Thus, its activity is structured around the following lines of action: (1) improvement of perception technologies, (2) optimisation of communication technologies, (3) redesign of occupant damage mitigation technologies and (4) development of new delivery solutions. This paper summarizes the most critical aspects of each of these lines and the considerations to be taken into account to ensure the implementation of autonomous and connected driving in complex environments. In doing so, it takes into account the particular conditions of urban environments and their implication for the deployment of autonomous mobility.
1 Introduction
Full self-driving automation is the next big mobility-related disruptive innovation not just for the technology development it implies, but also for the benefits it brings to the people and goods transport [1]. In the midst of the transition to decarbonisation, the autonomous and connected driving has as much potential to reduce energy consumption, through more efficient driving, as it does driving stress because it does not require a driver. Another potential benefit is related to cost reductions in the logistics sector and shorter delivery times.
From the societal point of view, fully autonomous and connected driving has potential to enhance road safety by eliminating human errors. So much so that [2] have estimated that driverless cars may reduce traffic fatalities by up to 90% by mid-century. However, seen from the technological side, there is still a long way to go before we see autonomous vehicles on the roads, especially in urban environments.
Anzeige
The complexity of urban environments, with numerous actors creating situations of enormous variability and technical complexity, raises many difficulties to deploy safe autonomous solutions. In this sense, Integra project, born of collaboration between four Spanish R&D centres, is focused on developing technological solution that enable a ultra-safe level 3–4 autonomous driving in complex urban environment, from a holistic approach, considering both passengers and goods transport. To this end, the project addresses the following interdisciplinary areas: autonomous driving functionalities, connectivity, occupant safety and autonomous goods delivery.
2 Autonomous Driving Functionalities
While the success of autonomous driving in urban environments depends on many factors, one of the most important falls on the autonomous driving functionalities have a correct and accurate perception of the environment. To achieve this, it is critical that perception algorithms have access to several groups of data sources. Specifically, to the following data sources: (i) environment real data provides by the integrated sensors; (ii) environment real data provided by external sensors and (iii) context annotated data provided by virtual reconstructions of the real world.
In the framework of INTEGRA project, work was done to exploit these three categories to the highest level in order to address the particular conditions of the urban environments. As a result, the following difficulties were identified:
Due to the topology of the cities, there are many regions of interest that can be occluded by buildings, vehicles or other items. Consequently, onboard sensors might be insufficient for detecting all kind of dangerous situations.
With respect to localization, urban scenarios quite often include spaces with low satellite coverage or with GNSS signals affected by multipath disturbances.
Perception obstacle detection algorithms must be very sensible to the presence of Vulnerable Road Users (VRUs) and robust against false positives.
To address them, the following solutions were considered:
A group of cooperative sensors, in order to complete the perception of vehicles.
A LiDAR-based positioning system to improve localization.
A specific module for classifying and precisely positioning the VRUs and vehicles in relative 3D coordinates by using a vision-based system.
Anzeige
2.1 Functions Considerations for the Algorithms
In parallel, the driving functions will have to take decisions on their own. It is therefore important that the SW architecture of the functions includes the “Path Planning” module, the “Control” module and the “Actuation” block.
After studying it through several use cases, it is critical that in urban environments the following workflow is followed:
The information from the surroundings of the vehicle, captured by sensors, is processed by the perception layer, which enters the “Path Planning” module. In this layer, all the information is organized and enriched according to the needs the “Behaviour Planning” block.
In the “Behaviour Planning” block, the information is used to execute functionalities such as lane change and speed adaptations.
Finally, the “Control” module calculates lateral and longitudinal action to be performed by the vehicle’s actuator actions according to the instructions of the previous block.
3 Connectivity
Complementing the above, the future of driving also relies on facilitating the communication between vehicles, especially in urban environments due to their complexity. The ability to receive information from other vehicles is essential to achieve safe driving. Thus, within Integra project, OMNeT + + and SUMO simulators were coupled in order to evaluate the enhancement in driving safety facilitated by Vehicle-to-Everything (V2X) communications.
After several simulations based on real rather than synthetic mobility data with different urban scenarios, it was observed that the most significant parameter for a safe urban autonomous driving is the position error perceived by the vehicles, due to the lack/loss of V2X messages. Briefly, this error is defined as the difference between the current position of one vehicle and the position known by its neighbors.
One example is show in Fig. 1, which plots the position error perceived by the vehicles involved in a collision scenario when they drive at 60 km/h. These values are presented in function of time, being the first error, a little before second 12. At this point, the distance between both vehicles is large and several messages are lost. Later on, nearly no message is lost.
Fig. 1.
Position error of the vehicles involved in the collision when driving at 60 km/h.
Subsequent larger errors are caused by the consecutive loss of messages, which normally occur when vehicles are apart and have buildings between them.
To obtain these results, a framework was developed that allows capture of real vehicle behavior from traffic video cameras. Graphically, Fig. 2 illustrates an example of the detection and tracking results of several combined traffic cameras from which information is obtained to feed the simulators.
Fig. 2.
Tracking results for the combination all cameras from the US 101 database [3].
In parallel to the described simulations, another parameter to take into account in terms of connectivity is location data. Presently, standalone GNSS systems are prone to errors exceeding 1 m due to interferences. These inaccuracies are particularly pronounced in urban canyons, where GNSS may produce errors in the tens of meters. To address this challenge the join use of 5G-NR technology and GNSS-based localization was explored within Integra project.
See in the Fig. 3 the performance of 5G-NR versus stand-alone GNSS in three simulation environments: urban, suburban, and rural.
Fig. 3.
CDF Position Error in three scenarios (rural, suburban and urban).
Note that GNSS performs considerably better than cellular systems in environments where there are a large number of visible satellites for users (rural and suburban). On the other hand, in urban environments, the error in position estimation through GNSS begins to increase due to the loss of visible satellites. In contrast, 5G-NR benefits from a high density of nodes, which results in a reduction of positioning error.
4 Occupant Safety
The deployment of autonomous driving in urban environments must go hand in hand with a number of readjustments of the occupant protection system. Unfortunately, the entry of autonomous driving does not guarantee the eradication of all accidents. Thus, it is important that in the event of accident the occupant is provided with an appropriate restraint system.
From this point view and with an eye on the progress of the Advanced Driving Assistance Systems (ADAS), the positions listed in Table 1 will be usual in the near future. Consequence of the greater freedom of movement that involves increasing the level of automation of driving, the occupants face a safety issue: the bad performance of the restraint systems due to different positions from the nominal one.
Table 1.
Autonomous driving scenarios in the near future.
A series of four sketches depicting a person in a car seat, illustrating different seating positions labeled as Relax, Entertainment, Work, and Rest. Each sketch shows the person in a blue outfit with a red head, adjusting posture and seat position according to the activity. The car interior is visible, providing context for the seating arrangements.
Integra project addresses this question by first analysing the risks associated with the positions described in Table 1 through crash simulations, and then developing and implementing an algorithm for monitoring the driver’s position.
Taking into account the results of the simulations at various crash speeds, it can be seen that the posture of the occupants is the variable which most influences on the correct performance of the protection elements. Consequently, seat-embedded solutions provide the best protection for the occupants, as their position relative to the person in the event of a crash is always the same.
However, in certain positions, even with seat-embedded solutions, simulations shows that restraint system cannot adapt to the occupant’s pose. In these cases the driver is required to modify his or her posture. Within Integra project, work was proceeded on a deep learning algorithm capable of predicting 33 key points on the driver’s body and from them estimating the driver’s pose, with a view to applying it to adaptive restraint systems in the near future.
5 Autonomous Goods Delivery
Finally, while occupant safety was mentioned above, autonomous driving also requires a number of actions in the field of goods transport in order to maximize safety. On this road, there are several important lines of action that range from guaranteeing the integrity of the goods to designing an optimal and safe vehicle that helps to reduce the delivery times. The following methodology has been used in the Integra project to address these challenges:
5.1 Technical Characterization of Urban Delivery Routes and Smart Mobile Locker Design
In order to know the vibrations that the goods are subjected to during the distribution, a total of 15 last-mile routes were monitored in four Spanish different cities. The study was carried out with the same type of delivery van and a data recorder (DR) device, to identify linear accelerations (vertical vibration). The Fig. 4 shows the maximum envelope curve that was obtained. Based on the severity of interurban routes (PSD) monitored, a smart mobile locker design was also developed (Fig. 5).
Fig. 4.
Envelope of maximum PSD curve obtained from the 15 routes under study.
In parallel, a vibration test was developed to simulate the reality of the monitored last mile distribution routes and to generate further synthetic data to support the design development.
6 Conclusions
To recap, the deployment of autonomous driving in urban environments is not imminent. In order for the autonomous driving to become a reality in cities, the appropriate technology must be first available and a safety level equal to or higher than the current one must be guaranteed. In this context, the success of smart mobility in urban environments lies in the following key aspects: the correct and accurate perception of the environment, the proper performance of V2X communications, the availability of an appropriate restraint system and the development of vehicles adapted to the characteristics of the environment. Working along these lines ensures the implementation of autonomous and connected mobility both in urban environments and in specific logistics environments, where the application of high safety systems is necessary.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Nastjuk, I., Herrenkind, B., Marrone, M., Brendel, A.B., Kolbe, L.M..: What drives the acceptance of autonomous driving? An investigation of acceptance factors from an end-user’s perspective. Technol. Forecast. Soc. Change 161 (2020)
2.
Bertoncello, M., Wee, D.: Ten Ways Autonomous Driving Could Redefine the Automotive World. McKinsey & Company (2015)
3.
U.S. Department of Transportation Federal Highway Administration.: Next Generation Simulation (NGSIM) Vehicle Trajectories and Supporting Data. ITS DataHub (2016)
