Back to datasets
Dataset assetOpen Source CommunityLast‑Mile DeliveryLogistics Data Analysis

Cainiao-AI/LaDe-D

LaDe is a public last-mile delivery dataset containing millions of parcels from industry. The dataset has three distinctive characteristics: (1) Large scale – it involves 21,000 couriers handling 10,677,000 parcels over six months. (2) Comprehensive information – it provides raw parcel details such as location and time requirements, as well as task event information that records the courier's location and time at task acceptance and completion. (3) Diversity – the dataset includes pick‑up and delivery data from multiple cities, each exhibiting unique spatio‑temporal patterns.

Source
hugging_face
Created
Nov 28, 2025
Updated
Jun 22, 2023
Signals
578 views
Availability
Linked source ready
Overview

Dataset description and usage context

Dataset Overview

Dataset Name: LaDe

License: Apache-2.0

Tags:

  • Spatial-Temporal
  • Graph
  • Logistic
  • Last-mile Delivery

Size Category: 10M < n < 100M

Dataset Features

Feature NameData Type
order_idint64
region_idint64
citystring
courier_idint64
lngfloat64
latfloat64
aoi_idint64
aoi_typeint64
accept_timestring
accept_gps_timestring
accept_gps_lngfloat64
accept_gps_latfloat64
delivery_timestring
delivery_gps_timestring
delivery_gps_lngfloat64
delivery_gps_latfloat64
dsint64

Dataset Splits

Split NameBytesSample Count
delivery_jl5,568,30931,415
delivery_cq168,574,531931,351
delivery_yt36,796,326206,431
delivery_sh267,095,5201,483,864
delivery_hz335,088,0001,861,600

Download Information

  • Download Size: 290,229,555 bytes
  • Dataset Size: 813,122,686 bytes

Dataset Description

LaDe is a public last‑mile delivery dataset containing millions of parcels from industry. The dataset has the following characteristics:

  1. Large Scale: It involves 10,677k parcels and 21k couriers, covering six months of real‑world operation.
  2. Comprehensive Information: It provides raw parcel details such as location and time requirements, as well as task event information that records the courier's location and time at task acceptance and completion.
  3. Diversity: The dataset contains data from various scenarios, such as parcel pick‑up and delivery across multiple cities, each with its own unique spatio‑temporal patterns.

Dataset Usage

When using this dataset for research, please cite the relevant paper: {xxx}

Need downstream help?

Pair the dataset with AI analysis and content workflows.

Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.

Explore AI studio