Surrogate Safety Measures Prediction at Multiple Timescales in V2P Conflicts Based on Gated Recurrent Unit

Matteo Miani, Matteo Dunnhofer, Christian Micheloni, Andrea Marini and Nicola Baldo
Additional contact information
Matteo Miani: Polytechnic Department of Engineering and Architecture (DPIA), University of Udine, Via del Cotonificio 114, 33100 Udine, Italy
Matteo Dunnhofer: Department of Mathematics, Computer Science and Physics (DMIF), University of Udine, Via delle Scienze 206, 33100 Udine, Italy
Christian Micheloni: Department of Mathematics, Computer Science and Physics (DMIF), University of Udine, Via delle Scienze 206, 33100 Udine, Italy
Andrea Marini: Department of Languages, Literatures, Communication, Education and Society (DILL), University of Udine, Via Margreth 3, 33100 Udine, Italy
Nicola Baldo: Polytechnic Department of Engineering and Architecture (DPIA), University of Udine, Via del Cotonificio 114, 33100 Udine, Italy

Sustainability, 2021, vol. 13, issue 17, 1-18

Abstract: Improving pedestrian safety at urban intersections requires intelligent systems that should not only understand the actual vehicle–pedestrian (V2P) interaction state but also proactively anticipate the event’s future severity pattern. This paper presents a Gated Recurrent Unit-based system that aims to predict, up to 3 s ahead in time, the severity level of V2P encounters, depending on the current scene representation drawn from on-board radars’ data. A car-driving simulator experiment has been designed to collect sequential mobility features on a cohort of 65 licensed university students who faced different V2P conflicts on a planned urban route. To accurately describe the pedestrian safety condition during the encounter process, a combination of surrogate safety indicators, namely TAdv (Time Advantage) and T 2 (Nearness of the Encroachment), are considered for modeling. Due to the nature of these indicators, multiple recurrent neural networks are trained to separately predict T 2 continuous values and TAdv categories. Afterwards, their predictions are exploited to label serious conflict interactions. As a comparison, an additional Gated Recurrent Unit (GRU) neural network is developed to directly predict the severity level of inner-city encounters. The latter neural model reaches the best performance on the test set, scoring a recall value of 0.899. Based on selected threshold values, the presented models can be used to label pedestrians near accident events and to enhance existing intelligent driving systems.

Keywords: ADAS; traffic safety; surrogate safety measures; driver behavior; Gated Recurrent Units; driving simulator (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2021
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2071-1050/13/17/9681/pdf (application/pdf)
https://www.mdpi.com/2071-1050/13/17/9681/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:13:y:2021:i:17:p:9681-:d:624041

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().