Outsight Introduces the First LiDAR Pre-processing Software Engine

LiDAR technology is a strong emerging trend in the Computer Vision landscape: over $3 billion have been invested in the Hardware aspect of LiDAR technology since 2019 and only the top five American LiDAR Hardware Manufacturers are expected to grow from less than $150 million sales in total in 2020 to over $4 billion in 2024.
However, effectively using LiDAR data in real-time is a complex, expensive and long endeavor even for the best 3D expert engineers. Assessing and selecting the right hardware among the profusion of dozens of manufacturers without any standard makes it even more challenging and threatens to slow down market adoption.
In that context, software for LiDAR is expected to account for at least 50% of the value and over 65% of gross profit across many markets: integrators and solution providers that are not 3D experts require a processing solution that solves the complexity of using LiDAR data regardless of the hardware supplier.

PARIS–(BUSINESS WIRE)–Outsight, the pioneer of 3D Spatial Intelligence solutions, announced today the launch of a game-changing product: a full Spatial Intelligence platform based on Lidar technology.

It’s the first real-time LiDAR Software Engine that allows application developers and integrators to seamlessly use LiDAR data from any hardware supplier.

Created as a turnkey solution, it enables leveraging 3D Spatial Intelligence’s unique value while avoiding the complexity of processing 3D data in real-time.

Being a LiDAR-agnostic solution, it saves the customer the hassle of assessing and choosing the most appropriate LiDAR for each application. LiDAR can now be easily used by everyone!

This new product offering follows thorough early customers’ validation processes across several different markets (Robotics, Automotive, ITS, Security & Surveillance), geographies (USA, Europe, Asia) and user profiles (Market-leading corporates, Start-ups and Universities) as well as strategic partnership agreements and collaborations with the most prominent LiDAR suppliers in the USA and Asia, including Velodyne (NASDAQ: VLDR), Ouster (NYSE: OUST), Hesai, and Robosense.

Whether it is for monitoring the flow of people or goods, builders of Mobile Robots & Vehicles as well as Integrators of Solutions are increasingly interested in leveraging the unique value of real-time 3D Spatial Intelligence that LiDAR technology creates, but don’t want to deal with the complexity of processing RAW LiDAR data.

Moreover, for the best professionals in most applications, going through the hassle of assessing, selecting and using the right LiDAR sensor out of dozens of hardware suppliers and more than a hundred available products, without any standard, is also a time-consuming, non-value added and inefficient use of engineering resources.

Turning any LiDAR into a Spatial Intelligence device

The solution is a real-time software engine that turns any LiDAR into a Spatial Intelligence device.

It overcomes the complexity of using RAW 3D data, so any application developer or integrator can efficiently use LiDAR in its own solutions without needing to become a 3D LiDAR expert.

It provides a comprehensive set of fundamental features that are commonly required in almost every application (e.g., Localization& Mapping, 3D SLAM, Object ID & Tracking, Segmentation & Classification, among others).

Because it only requires an ARM CPU and its AI doesn’t rely on Machine Learning, the solution is power-efficient and doesn’t need any Training or Annotation efforts.

According to Raul Bravo, President and co-founder of Outsight: “The hardware aspect of LiDAR is becoming a commodity with prices decreasing very quickly together with impressive performance improvements.

However, this new animal in the Computer Vision landscape remains a complex technology for most customers to use. As it happened every time in modern-day History of technology adoption, we’re convinced that the key condition required for LiDAR to become mainstream is the emergence of enabling software.”

LiDAR-agnostic

There is no LiDAR hardware that can fit all applications and contexts: the company has built an enabling computing layer, performing all required task regardless of the end-user application and LiDAR supplier, so integrators and solution providers are not constrained by the limitations of specific sensors.

A new standard adopted by the leaders of the market

The company has seduced not only leading customers in fields such as Smart City, Robotics and the Automotive industry, but has also established strategic partnerships with the world’s most renowned LiDARs manufacturers, such as American leaders Velodyne (NASDAQ: VLDR) and Ouster (NYSE: OUST), or the Chinese Hesai and Robosense.

The launch also follows the successful deployment of Outsight’s technology at Paris Charles de Gaulle airport of the ADP group, to provide accurate real-time monitoring of people flow while preserving private data.

Outsight has grown rapidly by integrating new features into its LiDAR-based processing solutions that enable systems to perceive, understand and interact with their surroundings in real time.

With a new generation of hardware and software pre-processing engine, connected to any LiDAR of the market, Outsight offers a unique level of simplicity, efficiency and versatility.

Award-Winning Technology

In less than a year, Outsight has successfully designed and industrialized this new generation of LiDAR processing solutions, which has been the subject of 63 patent applications.

Outsight’s innovation won many awards, including the prestigious Best of CES Innovation Award in Las Vegas, and it’s the youngest company ever to have won the Prism Award by the world leaders in photonics.

About Outsight

Outsight’s Physical AI continuously digitizes the movement, behavior, and interactions of people, vehicles and robots at scale, producing real-time Spatial Intelligence built on 3D LiDAR data and enriched with other available data sources.

Operators of large, complex environments, including transportation hubs such as airports and train stations, sports and entertainment venues, road infrastructure, and industrial sites, rely on this Spatial Intelligence to enhance operations, improve visitor experiences, and strengthen safety and security.

We support customers across five continents from our offices in Paris, San Francisco, Singapore and Hong Kong, driven by the belief that accelerating the global adoption of Spatial Intelligence will lead to a smarter, safer, and more sustainable world.

Frequently Asked Questions

What does a LiDAR preprocessor actually output, and what format does it use?

A LiDAR preprocessor converts raw point clouds (millions of x, y, z measurements per second per sensor) into structured streams of tracked, classified entities: people, vehicles, and robots, each carrying a unique anonymous ID, a 3D bounding box, a classification label, and a behavior tag. Outsight's Augmented LiDAR Box emits this in OSEF (Outsight Spatial Exchange Format), an open format covering per-frame point clouds, tracked entity arrays, and classification metadata. This structured output feeds directly into the SHIFT platform, enabling situational awareness, analytics, and open integrations without requiring downstream tools to handle raw sensor data. The anonymity is built into the process, since LiDAR geometry captures shape and motion rather than faces or license plates.

Why doesn't LiDAR preprocessing require GPU hardware or machine learning training data?

Classical LiDAR preprocessing pipelines rely on geometric algorithms rather than neural networks trained on annotated image datasets. Because the processing is geometry-based, it runs on an ARM CPU without a dedicated GPU, keeping the hardware footprint small. Outsight's Augmented LiDAR Box applies exactly this principle: its real-time software engine eliminates the annotation and retraining cycles that camera-based computer vision systems require when they encounter new environments or lighting conditions. The trade-off inherent to the geometry-based approach is that it works on shape and motion rather than appearance, which is also why it cannot capture faces or license plates, making the pipeline anonymous by definition.

How many LiDAR sensor models is a hardware-agnostic software platform expected to support?

Hardware fragmentation is a core challenge in the LiDAR market: dozens of manufacturers produce sensors with different data formats, coordinate systems, and point-cloud densities, and no universal standard exists. Outsight's SHIFT platform currently supports 210+ compatible LiDAR models from manufacturers including Hesai, RoboSense, Ouster, and Seyond. That breadth matters for operators because it preserves the ability to mix sensor brands across a single site, switch suppliers without rewriting the processing stack, and optimize per-zone cost without vendor lock-in. In practice, this multi-vendor compatibility is what allows deployments like Dallas Fort Worth, the world's largest 3D LiDAR airport installation, to scale across a complex site without being constrained to a single hardware supplier.

Can the same LiDAR processing software run on both mobile robots and fixed infrastructure sensors?

Yes. The core algorithmic functions, localization, mapping, SLAM, object detection, tracking, and classification, are sensor-position-agnostic: the pipeline processes point clouds regardless of whether the emitting sensor is mounted on a ceiling, a pole, an autonomous mobile robot, or a vehicle. Fixed-infrastructure deployments and robot-borne deployments differ in coverage geometry and latency requirements, so configuration tuning differs, but the underlying processing engine is shared. Outsight's Augmented LiDAR Box was built on this principle, validating the same software base across robotics, ITS, and security use cases. In production infrastructure deployments, such as airports and train stations, Outsight runs this pipeline end-to-end in under 50 milliseconds, demonstrating that a single processing engine can serve both mobile and fixed sensor configurations at operational scale.

What happened at Paris-CDG airport that proved the technology was production-ready before the product launch?

Before the formal product announcement, Outsight completed a live deployment at Paris Charles de Gaulle Airport under Groupe ADP. The deployment provided real-time people-flow monitoring across the terminal while preserving passenger privacy by design: because LiDAR captures geometry rather than images, no biometric or facial data was recorded. That live airport environment, with its high passenger density, variable lighting, complex geometry, and strict data-protection obligations, served as the production validation for what would become the SHIFT platform. Successfully operating under those conditions preceded both the commercial launch of the Augmented LiDAR Box and the subsequent partnerships with sensor manufacturers.