How Gaode Maps Built a Real‑Time Lakehouse for Billion‑Scale Trajectory Data

This article details Gaode Maps' end‑to‑end lakehouse solution for massive, high‑frequency trajectory data, covering the challenges of real‑time visibility, query performance, and storage cost, and explaining how a hot‑warm‑cold tiering architecture built on Apache Flink, Paimon, StarRocks, Redis and Lindorm delivers millisecond‑level queries while cutting storage expenses.

Apache FlinkApache PaimonData Tiering

0 likes · 19 min read

How Gaode Maps Built a Real‑Time Lakehouse for Billion‑Scale Trajectory Data

JD Tech

Jun 12, 2021 · Artificial Intelligence

DeepDualMapper: A Gated Fusion Network for Automatic Map Extraction Using Aerial Images and Trajectories

The paper presents DeepDualMapper, a gated‑fusion deep network that combines aerial imagery and vehicle trajectory data to automatically generate high‑precision maps, detailing its architecture, gated and refinement modules, and experimental validation on three city datasets.

Deep LearningTrajectory DataU-Net

0 likes · 7 min read

DeepDualMapper: A Gated Fusion Network for Automatic Map Extraction Using Aerial Images and Trajectories

JD Tech Talk

Nov 9, 2020 · Big Data

Trajectory-Based Population Flow Analysis for COVID‑19 Prevention Using HBase and Spark

The article presents a comprehensive big‑data solution that stores massive GPS trajectory records in HBase, processes them with Spark to identify individuals who visited a pandemic source region, and visualizes their spatio‑temporal distribution in target cities to support precise epidemic control measures.

Big DataCOVID-19HBase

0 likes · 8 min read

Trajectory-Based Population Flow Analysis for COVID‑19 Prevention Using HBase and Spark