labs

KOREAN

NAVER LABS’



On-road

Intelligence

NAVER LABS is conducting ongoing research on
roadway autonomous
driving technology by producing

machine-readable HD maps and
updating them after sensing
changes on road surfaces.

Such maps are able to carry out
precise localization even in complex
metropolitan settings.

On-road Intelligence is an integrated solution
for roadway
autonomous driving is the first in the IT
industry to have

received approval for provisional autonomous driving operations
by
the Ministry of
Land, Infrastructure and Transport, and
will be verified
and enhanced in a real road environment.

NAVER LABS is conducting ongoing research on roadway autonomous driving technology by producing machine-readable HD maps and updating them after sensing changes on road surfaces. Such maps are able to carry out precise localization even in complex metropolitan settings. On-road Intelligence is an integrated solution for roadway autonomous driving is the first in the IT industry to have received approval for provisional autonomous driving operations by the Ministry of Land, Infrastructure and Transport, and will be verified and enhanced in a real road environment.

On-road Intelligence
Platform

On-road machine platform for autonomous driving machines on the road

We are advancing autonomous driving technology in all areas including on-road localization, perception,
planning and control to produce HD maps for

roadway autonomous driving using hybrid HD mapping and
the ACROSS solution. We will integrate these technologies and data to later develop


a customizable roadway autonomous driving machine platform for various purposes
such as logistics, delivery, and unmanned stores.

Key technology of autonomous driving #1

Hybrid HD

Mapping

The new solution to creating precise maps for
autonomous driving machines

The Hybrid HD mapping technology is NAVER LABS’s own
HD
map production format that organically
combines
large-scale
regional aerial photographs in city units with
the data
collected by the in-house
developed R1 mobile
mapping
system (MMS). Compared to HD maps made by
the existing
MMS vehicle,
it can maintain a high degree of
accuracy even
with a major reduction in cost and period
of production.

Relevant patents
  • 1) Method for the convergence and optimization of aerial photo-based maps and ground-level MMS maps for the production of precise road maps
  • 2) Aerial video convergence technology and device for extraction of road information
Relevant patents
  • 1) Method for the convergence and optimization of aerial photo-based maps and ground-level MMS maps for the production of precise road maps
  • 2) Aerial video convergence technology and device for extraction of road information
Key technology of autonomous driving #2

HD Map


Update Solution,



ACROSS Project

Technology that keeps the HD maps for
autonomous machines up to date

It is important for maps to be up to date of the rapidly

changing urban environments.
ACROSS is the HD map

updating solution for keeping the information up to
date
on the continuously
changing road environments.

By equipping a large number of ride-sharing vehicles
with
mapping devices
comprising low-cost sensors,

we can simultaneously confirm and update multiple
changes
in the road data
of wide urban areas.

Key technology of autonomous driving #3

Localization


on the road

Precise localization technology, seamless even in
urban areas with weak signals

It is important to accurately ascertain the current location
in
roadway autonomous driving situations.
The
localization
technology of NAVER LABS utilizes HD maps
developed
in-house, such as a virtual sensor,
and because
it matches
the data obtained from diverse sensors
including the existing
LiDAR, Cameras, IMU,
and Wheel
Encoders, it is capable of
accurately and stably estimating
the current location on
the road even
in areas where GPS
signals are weakest
such as complex urban areas and
tunnels.

Relevant patents
  • 1) Multiple-sensor technology for location estimation of outdoor self-driving vehicle
  • 2) Encoder frame device and vehicle positioning system using said device
Relevant patents
  • 1) Multiple-sensor technology for location estimation of outdoor self-driving vehicle
  • 2) Encoder frame device and vehicle positioning system using said device
Key technology of autonomous driving #4

Perception & Prediction

Using diverse sensors such as LiDAR, Cameras and Radar along with

deep learning technology, road
information such as lane lines near the
vehicle,
traffic lights, or traffic signs is continuously recognized in all

directions up to
360º to predict paths of motion of nearby vehicles and
pedestrians.

Key technology of autonomous driving #5

Planning & Control

Based on precise localization of the current position and the result of
recognition
of the surrounding
environment, nearby motions are predicted
and the pathway
to the destination is generated in real-time. In a
roadway
autonomous driving
situation, stable following of the path of motion is
enabled by reliable steering/

speed control.

Relevant patents
  • 1) Camera-based lane change timing prediction system and data collection method for system learning
  • 2) Countdown notification for the availability of lane change

Challenge to Change

How might autonomous driving technology change the way we live?

A place for information and services, moving independently and making new connections;


we introduce to you an image of a future city formed by the technology of NAVER LABS.