The localization stackfor robots that can't fail.

Ports. Tunnels. Mines. Race circuits at 200 km/h. Exwayz delivers <2 cm localization in every environment.

Trusted in production by

Technology

State of the art 3D LiDAR algorithm

A new standard in 3D LiDAR SLAM robustness. Our proprietary algorithm has been built to overcome major edge cases faced by robotics engineers.

LiDAR only & agnostic

Compatible with any LiDAR on the market. No camera, no GPS, no IMU required.

Platform agnostic

Any robot, any form factor. If it moves and carries a LiDAR, Exwayz runs on it.

Environment agnostic

Built for real-world complexity. Operates continuously across any environment, without interruption.

Products

Meet Exwayz Engine.

01 Exwayz Mapping

Stand alone product to build centimeter-accurate large and dense point clouds with raw LiDAR data to improve field operations based on ground data.

Loop closure Multi session Georeferenced maps

02 Exwayz Navigation

Map-based localization working in real time and running on the embedded computer giving accurate position of the LiDAR (1 cm and 0.1°).

CPU only Edge cases by design No other sensor needed

03 Exwayz Perception Ground truth & real-time perception · LiDAR-native · Semantic scene understanding

Request early access →

<2 cm

Position accuracy

0.1 °

Orientation error

20+ Hz

Update rate

<15 ms

Latency per frame

Case Studies

Precision localization, in the real world

Port of Rotterdam, the Abu Dhabi circuit, La Défense.
Here’s what happens when you stop wrestling with localization and start shipping.

Heavy-duty logistics

Embotech: Bringing autonomy to port logistics

Embotech's Autonomous Terminal Tractors operate 24/7 across container yards with poor GNSS coverage and featureless terrain. Exwayz delivers centimeter-level accuracy amid constantly moving cranes, stacks and vehicles.

Read story →

Autonomous racing

FR4IAV: LiDAR localization at 250 km/h

The first French A2RL team runs Exwayz Navigation on a fully autonomous race car. Three LiDARs fused in real-time, GNSS-independent, centimeter-level precision at 250+ km/h on the Abu Dhabi circuit.

Read story →

Geospatial mapping

Cyvl.ai: Automating road condition assessment

200,000+ km of US roads mapped with centimeter-level geo-referencing. Exwayz SLAM handles multi-hour surveys, urban canyon signal loss, overlapping trajectories and real-time dynamic object removal.

Read story →

View all case studies →

Compatibility

Works with your hardware, today

Exwayz is sensor-agnostic and platform-independent. If you use any of these LiDARs or platforms, you’re ready to integrate.
No hardware changes required.

LiDAR Sensors

Ouster

Hesai

Velodyne

RoboSense

Livox

Innoviz

Seyond

Ouster

Hesai

Velodyne

RoboSense

Livox

Innoviz

Seyond

Ouster

Hesai

Velodyne

RoboSense

Livox

Innoviz

Seyond

Ouster

Hesai

Velodyne

RoboSense

Livox

Innoviz

Seyond

Compute & OS

Ubuntu Linux

20.04/22.04/24.04

ROS 1 & ROS 2

ROS package shipment

Lightweight resources

Jetson Nano, Rubik Pi 3

CPU-only

No GPU required

RTMaps

Timestamp, sync. multi sensors

Industries

Set for any application, any environment

From mines to military patrols: if it moves autonomously and needs to know where it is, Exwayz handles it.

Port & Heavy Logistics

GNSS denied Dynamic objects Changing environment

Warehouse & AMRs

GNSS denied Changing environment Repetitive aisles

Autonomous Shuttles

Urban canyons Dynamic objects Vibrations

Mining & Defense

GNSS denied Indoor Featureless

Patrols & Inspections

Multi levels Indoor In/outdoor

Autonomous Racing

Vibrations Sensor fusion High velocity

How it works

From first contact to full deployment

Three phases. No commitment until you see it work on your own hardware.

Phase 01

Demo

Send us back a ROSbag. We process your data and build a localization demo in hours, using your exact sensor setup and environment. See the results before writing a single line of code.

Phase 02

Free trial

Test the full software on your own machine or robot for two weeks. No credit card, no infrastructure changes. Your hardware, your environment, your team, full features.

Phase 03

Scale together

Per-robot, per-year licensing that grows with your fleet. Volume pricing available from day one. We stay close as you scale, from pilot to full production deployment.

Launchpad

A 3-month program for 5 teams building ambitious mobile robotics projects in challenging environments.
We provide engineering advice, LiDAR loans, and free licenses.

Early accessNew features before release

Engineering supportDirect line to our team

Peer networkConnect with builders in production

ResourcesSDK docs, datasets, tutorials

Academic

Making research move forward

Free SDK access for universities and research labs working on SLAM, autonomous navigation, and mobile robotics.
Datasets, engineering support, and a direct line to our team.

About