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*).
Dieses Kapitel untersucht die Integration von eCalls in das Incident Management System (IMS) des Straßenbetreibers, um dessen Auswirkungen auf die Reaktionszeiten nach einem Unfall zu bewerten. Im Mittelpunkt der Studie steht der Abgleich von eCalls mit den Störungsdaten des österreichischen Autobahnbetreibers ASFINAG, wobei die Messgröße "eCall-Gain" eingeführt wurde, um potenzielle Zeitersparnisse zu messen. Zu den wichtigsten Ergebnissen zählen signifikante Verkürzungen der Reaktionszeiten, wobei automatische eCalls eine größere Effizienz aufweisen als manuelle. Das Kapitel diskutiert auch die Herausforderungen bei der genauen Beurteilung der gesamten Ereignismanagementkette und schlägt zukünftige Forschungsrichtungen vor, wie etwa den Einsatz von C-ITS Communication zur präzisen Zeitstempelung. Die Ergebnisse unterstreichen das Potenzial von eCalls zur Verbesserung der Straßenverkehrssicherheit durch rechtzeitige Unfallwarnungen und die Verbesserung des gesamten Ereignismanagementprozesses.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
This study investigates the potential integration of eCalls, mandated by the eCall Legislation for vehicles from March 31, 2018, into road operator Incident Management Systems (IMS). We utilize eCall data from the “SRTI Ecosystem” and match it with IMS incidents to verify the eCall data and enhance the IMS with the most accurate available timestamp for accident occurrence. To assess the potential temporal gains of eCall integration, we introduce a metric to quantify the time saved by incorporating eCalls into event management systems. The proposed metric was evaluated over a three-month period in 2023, and the results indicate that integrating eCalls into IMS is a viable step to expedite the incident management process.
1 Introduction
To enable the timeliest driver and emergency service alerts following vehicle crashes, the so-called eCall Legislation was introduced [1]. Consequently, vehicles with type approval from March 31st, 2018, onwards must be equipped with an eCall system, which utilizes onboard sensors and communication technology to detect a crash and send an automated signal that includes the vehicles position and the time of the accident to a predefined emergency response center.
This paper aims to use eCalls by matching each event to corresponding incidents in a road operator Incident Management System (IMS), to verify the eCalls and to subsequently evaluate the potential of an integration of eCalls into the IMS.
Anzeige
To achieve this the authors have used eCall data from the “SRTI Ecosystem” (SRTI – safety related traffic information, DFRS – Data for Road Safety [2]), an EU-initiated partnership of EU transport ministries, Service Providers of traffic data and the automotive industry to pool SRTI Data for common use and free of charge. Every eCall is matched to the corresponding entries in the incident management system of Austria’s highway operator ASFINAG. Currently, certain key temporal parameters are either unknown or not available in the IMS, including the exact time when:
an accident occurs or
emergency vehicles are dispatched by the operator.
However, one consistent piece of information at our disposal is the timestamp of insertion into the IMS. The transmission of eCalls allows us to benchmark the performance of the event management chain, by measuring the delay between the time of the actual accident and the event was reported to and verified by the road operator staff. Since every minute is crucial in post-crash-responses, the potential for speeding up the process using automatically or manually triggered eCalls is investigated in this survey. We consider data limited to the Austrian highway network in the period from June to September 2023.
2 Related Work
The Fatality Analysis Reporting System (FARS) [3] is a comprehensive database that collects and maintains data on fatal traffic accidents within the United States, dating back to 1975. FARS data is extensively used to study the relationship between emergency response times and traffic accident outcomes, such as mortality and severity of injuries.
In [4], they investigate the evolution of emergency response times for highway accidents from 1975 to 2017, utilizing data from FARS. Their findings suggest a significant 50% improvement in response times, highlighting the potential impact of reduced response times on highway accident outcomes.
Anzeige
Utilizing the 2015 version of the FARS dataset, the study [5], found that the relationship between emergency medical service (EMS) response time and fatality odds is non-monotonic, with two critical thresholds: 5.5 min, the fastest decline in the chance of survival, and 17 min, the “gold time” for operating rescues. This suggests that accurately assessing the urgency of different crash scenarios is important for understanding the impact of response times on survival.
Similarly, the European baseline report defines several key performance indicators (KPIs) for road operators, one of which is post-crash care (i.e., the time elapsed between the emergency call and the arrival of emergency services, to the 95th percentile (P95)). In 2019, the post-crash-care KPI for Austrian highways was 28 min and 54 s [6].
The critical time thresholds pose an ambitious but desirable temporal target for minimizing the death toll and increasing the effectiveness of post-crash care. Because the post-crash-care KPI is widely used to assess emergency service response times, we incorporate a variation of this metric in our analysis, with the long-term goal of benchmarking the entire event management chain from accident to arrival of road operator vehicles at the accident location. At this point in time only the part of the incident management chain is investigated between the accident and the first entry into the event management system of a road operator.
3 Methodology
The development of the matching script was guided by the datascience process. Understanding the data was not just a matter of exploratory data analysis but required a thorough investigation of the inner workings of the entire event management chain, achieved by frequent discussions with domain experts. This led to the following steps in the processing pipeline, summarized in Table 1 (Fig. 1).
Table 1.
Data sources and preprocessing required for matching.
While the eCalls consist of event singular in location and datetime, the road operator events stretch over periods of time and road segments. A match occurs if the time and space of events from both sources overlap, with respect to a margin window (deltas).
Due to the inaccuracy in GPS positioning, the direction of the eCalls cannot be reliably map-matched to the graph representation of the highway network, prohibiting us from including the following matching rule:
To assess potential temporal gain by incorporating eCalls into event management systems, we introduce the ‘eCall-Gain’ metric, which adjusts the post-crash KPI defined in [7] to align with our specific use case. We define it as:
‘The time elapsed between the eCall following a collision and the timestamp of the first entry into the event management system of a road operator. The measurement unit is in minutes and seconds, calculated to the following percentiles: 25th, 50th, 75th, 85th and 95th, over all matched events.’
The timestamp tEVIS is the timestamp of the insertion into the IMS, teCall refers to the reported timestamp of the eCall. We denote the sorted data set of all temporal gains as \({\Delta }_{t}\) with N as the total number of matched events:
The Quantile Q(q) for a given percentile q is calculated finding the Index I and in case of non-integer value interpolating between adjacent datapoints. m denotes the integer and \(\alpha \) the fractional part. The following quartiles are calculated \(\{\text{0.25,0.50,0.75,0.85,0.95}\}\):
The eCall-Gain for automatic (26 tuples) and manual eCalls (10 tuples) and an average of both was calculated for all proposed percentiles, using linear interpolation with the percentiles that lie between two data points. The results can be found in Table 2. It has to be noted that in our used dataset, the eCalls are only available from a single car manufacturer, and while higher sample sizes and longer observation intervals are required for strong statistics, it has to be reiterated, that behind each of the counts in Fig. 2. Gain in minutes of DFRS over IMS, duplicates and outliers eliminated in DFRS, separated into automatic and manual eCalls, is an eCall, for which we found a corresponding entry in the IMS. Thereby the accident was verified, which implies that knowledge about the accident could have been obtained minutes earlier, which is highly relevant in case of especially heavy accidents.
Table 2.
Results for all eCalls (total average, manual and automatic)
Percentile
0.25
0.50
0.75
0.85
0.95
eCall-Gain avg
9 min36 s
12 min 46 s
16 min 49 s
19 min01 s
25 min 42 s
eCall-Gain automatic
10 min 35 s
13 min 54 s
16 min 43 s
17 min 5 s
23 min 25 s
eCall-Gain manual
7 min 23 s
10 min 37 s
18 min 40 s
23 min 32 s
26 min 17 s
Fig. 2.
Gain in minutes of DFRS over IMS, duplicates and outliers eliminated in DFRS, separated into automatic and manual eCalls, a Gain of 0 would refer to the two independent systems being synchronized.
Our understanding of the rest of the event management chain, namely the distribution of the differences of timestamps in the event management system and the timestamp of arrival at the accident location on the other hand is still lacking. Interviews with domain experts have raised several concerns. Frequently, emergency vehicles are dispatched or even arrive at the accident location before the entry is recorded in the road operator’s database. To address this issue, an evaluation of the distribution or precise arrival times at accident scenes can be obtained using alternative data sources, which were not available at the time of the conduction of our investigation. Further refinement of the eCall gain metric will be explored by employing technologies such as C-ITS Communication to obtain precise timestamps for the arrival of road operator vehicles at accident locations. Preliminary exploratory analysis in this field has produced promising results, and we plan to conduct future research in this area, to better understand and then benchmark the ent+ire incident management chain.
5 Conclusion
In conclusion, our approach to matching eCalls with accident data in Incident Management Systems has allowed us to evaluate our proposed metric, known as eCall gain. We have also found that our verification and benchmarking processes are applicable.
Looking ahead, we plan to expand our analysis by considering longer time intervals, with our current results serving as a promising proof of concept. These findings are well-positioned to contribute to the integration of eCalls into the Incident Management Systems (IMS) more effectively. By bridging the gap between eCalls and accident data, our work has the potential to enhance eCalls by enriching it with informations about impacts of an accident that is available in an IMS such as lane closures, traffic jam sizes, influence on travel times, and to improve the event management chain by reducing response times in a digitally connected world, in order to enhance road safety.
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.
Cruz, M.C., Ferenchak, N.N.: Emergency response times for fatal motor vehicle crashes, 1975–2017. Transp. Res. Rec. 2674(8), 504–510 (2020)CrossRef
5.
Ma, L., Zhang, H., Yan, X.: Smooth associations between the emergency medical services response time and the risk of death in road traffic crashes. J. Transp. Health 12, 379–391 (2019)CrossRef
6.
Nina, N.: Baseline report on the KPI Post-crash care. Baseline project. Vias Institute, Brussels (2022)
7.
Van den Berghe, W., et al.: Methodological guidelines – KPI Post-crash Care. Baseline project. Vias Institute, Brussels (2021)
