English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Estimated time of arrival (ETA) prediction, also known as travel timeestimation, is a fundamental task for a wide range of intelligenttransportation applications, such as navigation, route planning, andride-hailing services. To accurately predict the travel time of a route, it isessential to take into account both contextual and predictive factors, such asspatial-temporal interaction, driving behavior, and traffic congestionpropagation inference. The ETA prediction models previously deployed at BaiduMaps have addressed the factors of spatial-temporal interaction (ConSTGAT) anddriving behavior (SSML). In this work, we focus on modeling traffic congestionpropagation patterns to improve ETA performance. Traffic congestion propagationpattern modeling is challenging, and it requires accounting for impact regionsover time and cumulative effect of delay variations over time caused by trafficevents on the road network. In this paper, we present a practicalindustrial-grade ETA prediction framework named DuETA. Specifically, weconstruct a congestion-sensitive graph based on the correlations of trafficpatterns, and we develop a route-aware graph transformer to directly learn thelong-distance correlations of the road segments. This design enables DuETA tocapture the interactions between the road segment pairs that are spatiallydistant but highly correlated with traffic conditions. Extensive experimentsare conducted on large-scale, real-world datasets collected from Baidu Maps.Experimental results show that ETA prediction can significantly benefit fromthe learned traffic congestion propagation patterns. In addition, DuETA hasalready been deployed in production at Baidu Maps, serving billions of requestsevery day. This demonstrates that DuETA is an industrial-grade and robustsolution for large-scale ETA prediction services.

DuETA: Traffic Congestion Propagation Pattern Modeling via Efficient Graph Learning for ETA Prediction at Baidu Maps