Skip to content
Insights
Outsight joints Intel and Microsoft for the Enterprise Edge

Outsight, Intel and Microsoft bring Dynamic World Models to the Enterprise Edge

Enterprises can now run real-time Physical AI and site-wide situational awareness for people and robots on standard CPU-based edge infrastructure.


Integration of Outsight’s Shift platform with Microsoft Azure Local powered by Intel® Xeon® 6 SoC to be showcased at Microsoft Ignite (November 17-20, 2026)

SAN FRANCISCO, California, November 17, 2026: Outsight, the pioneer and global leader in Infrastructure-based Physical AI, recently named a leading vendor for Ambient Intelligence in the Gartner Hype Cycle for Physical AI, 2026, is expanding its technology collaboration with Intel to accelerate the deployment of Physical AI for real-time monitoring of people and object flows at enterprise scale. Outsight’s platform already equips some of the world’s largest international airports, train stations, hospitals and venues.

This combined architecture delivers turnkey Physical AI capabilities directly at the edge of enterprise networks.

A live demonstration of the integrated solution will be showcased at the Intel-hosted Dimes Gallery venue near Moscone Center in San Francisco during Microsoft Ignite, November 17-20, 2026.

Bringing Physical AI to enterprise infrastructure

The collaboration answers growing demand from operators of airports, train stations, industrial facilities, smart cities and large venues for real-time, privacy-preserving intelligence about how people, robots and vehicles move through physical environments.

Outsight’s Physical AI software platform, based on 3D LiDAR, continuously tracks the movement of people, vehicles, objects and robots. LiDAR data is fused with data from multiple other sources, including cameras, WiFi, boarding pass information and business systems such as AODB or POS.

Outsight’s Dynamic World Models process this data, and Motional Digital Twins deliver the results. These are real-time 3D replicas of physical flows, and they give end users a single operational interface to monitor, analyze, predict and optimize their environments.

Running natively on energy-efficient Intel® Xeon® 6 processors, the solution delivers cost-effective, scalable 3D AI inference at the edge.

Raising the bar: global situational awareness for robots

The integrated solution extends beyond people and vehicle flow monitoring to deliver global situational awareness: a shared, real-time representation of reality with temporal and spatial continuity, available to every operator and mobile robot on site.

In complex environments such as factories, warehouses and airports, mobile robots see only what their onboard sensors and processing allow. Blind spots, occlusions and instantaneous perception limit their understanding to their immediate field of view.

Dynamic World Models solve this with an infrastructure-level intelligence layer that continuously tracks every person, vehicle and object across the entire premises, each with a unique anonymous ID maintained over time.

This enables three foundational capabilities for robotics and operations:

  • Continuous Individual Tracking: Precisely positions each person or object across large distances, beyond visual line of sight, and makes every position instantly available to all robots and operators on site.
  • Situational Awareness: Perceives the environment globally, beyond the line of sight of each individual robot, and associates external data, event history and attributes with each tracked entity.
  • Actionable Insights and Prediction: Interprets situations at scale, derives decision-ready intelligence and anticipates future states across the entire premises.

Live demonstration at Microsoft Ignite

At Microsoft Ignite, Intel and Outsight will present a live, full-scale demonstration of the integrated Physical AI solution at an Intel-hosted venue.

Outsight joins Intel at Microsoft Ignite 2026

Attendees will see firsthand how the platform processes millions of 3D data points per second to detect, classify and track individuals and objects in real time, using anonymous, privacy-preserving 3D sensing.

The demonstration will feature real-time crowd density monitoring, queue management analytics, flow visualization through a live Motional Digital Twin, and predictive operational insights, all running on Intel-powered edge infrastructure within the Microsoft Azure Local framework.

Industry perspectives

“This collaboration is a major milestone in bringing Dynamic World Models and Infrastructure-based Physical AI to scale across airports, rail stations, factories and other critical sites worldwide,” said Raul Bravo, President Founder, Outsight.

“By optimizing our Shift platform for Intel® Xeon® 6 processors and integrating with Microsoft Azure Local, we are making high-performance, real-time Spatial Intelligence accessible to every enterprise operator, on standard edge infrastructure.”

“Enterprise AI is about bringing intelligence closer to where decisions are made, at the edge,” said Bhupesh Agrawal, General Manager, Enterprise Edge AI & Connectivity at Intel.

“Working with Outsight demonstrates how Intel’s edge computing platform, combined with best-in-class physical AI software and Microsoft Azure Local, enables organizations to transform their physical operations with real-time, CPU-based 3D AI inference. This is exactly the kind of scalable, cost-effective solution that enterprises need to unlock the full potential of physical AI.”

Key benefits of the integrated solution

  • Real-time Physical AI at the Edge: Outsight’s Physical AI processes massive 3D LiDAR data streams in real time on Intel® Xeon® 6 processors, delivering insights in under 50 milliseconds, entirely at the edge.
  • Privacy by Design: The solution relies on anonymous 3D data, fully compliant with GDPR and other strict privacy regulations.
  • Global Situational Awareness for Robots: Dynamic World Models give mobile robots premises-wide, context-aware intelligence that extends beyond their onboard sensors, enabling safer navigation, smarter path planning and coordinated fleet operations.
  • Scalable and Cost-Effective: CPU-based AI inference lowers total cost of ownership and energy consumption, enabling deployment across thousands of locations.
  • Hardware-Agnostic 3D Sensing: The Shift platform supports more than 210 LiDAR models from all leading manufacturers, giving operators the freedom to choose the best hardware for each environment.
  • Multi-Modal Data Fusion: LiDAR, cameras, WiFi, business data systems and other sources are fused into a unified spatial reference, delivering richer situational awareness than any single sensor.
  • Cloud-Managed Edge Deployment: Integration with Microsoft Azure Local provides centralized management, automated updates and enterprise-grade security for distributed edge deployments.

Proven at scale across five continents

Outsight’s Spatial Intelligence solutions are deployed at scale in major airports, train stations, stadiums, industrial sites and smart cities across five continents. More than 280 million annual passengers travel under the real-time measurement of its Spatial Intelligence Platform.

Outsight crosses 280 million passengers under real-time measurement, from the busiest US hubs to Singapore Changi

More than 280 million annual passengers now travel under real-time measurement by Outsight’s Spatial Intelligence Platform, from the busiest US hubs to Singapore Changi.

Read article →

The company has been cited seven times by Gartner, most recently as a leading vendor for Ambient Intelligence in the Gartner Hype Cycle for Physical AI, 2026, as the only 3D-native software company cited in this category. Gartner has also recognized Outsight as an emerging leader in the Digital Twin and Spatial Computing markets.

Gartner® Highlights Outsight as Digital Twin leader

Recognized for pioneering real-time digitalization of dynamic environments, Outsight sets itself apart by focusing on the live monitoring of people and vehicle flows.

Read article →

Outsight has received numerous industry awards, including the CES Best of Innovation Award, the Frost & Sullivan Global Technology Innovation Leadership Award, the Edge AI and Vision Product of the Year, the LiDAR Leader Award and the Airport Technology Excellence Award.


Related Articles

NEWS

Selected for French Tech 2030

The French State recognizes infrastructure-based Physical AI as a strategic technology and backs Outsight's next stage of growth worldwide.

NEWS

Outsight recognised among top leading AI companies

Outsight is included in the 2026 Mapping of AI Startups in France, a national initiative conducted with the support of Sopra Steria Ventures.

Let's connect

Send us a Message

Drop your email and we'll get back to you as soon as possible.

Frequently Asked Questions

  • What is Intel AMX and why does it matter for edge AI inference?

    Intel Advanced Matrix Extensions (AMX) is a hardware acceleration block built into Intel Xeon 6 processors that speeds up matrix multiplication, the core arithmetic operation in neural network inference. Because AMX runs on the CPU die rather than a discrete GPU or dedicated AI accelerator, it allows high-throughput AI workloads to run on standard server-class hardware without additional cards. For edge deployments where rack space, power budgets and procurement complexity are constraints, this matters: operators can run real-time 3D perception like Outsight's Shift Platform on the same compute they already use for other enterprise workloads.

  • How is Microsoft Azure Local different from a regular Azure cloud deployment?

    Microsoft Azure Local is a hyperconverged infrastructure product that runs Azure-consistent services on hardware physically located inside a customer's facility or at the network edge, rather than in a Microsoft datacenter. Workloads run on-premises but are managed through the same Azure control plane used for cloud services, including centralized policy, monitoring and update pipelines. For latency-sensitive Physical AI workloads like real-time 3D tracking from Outsight, this matters because the data never leaves the site: processing happens locally in under 50 milliseconds, while lifecycle management stays cloud-coordinated.

  • What is a Dynamic World Model and how does it differ from a Motional Digital Twin?

    A Dynamic World Model is the intelligence layer that continuously processes raw sensor data from across a site, assigning each detected person, vehicle or object a persistent anonymous ID and maintaining its position, trajectory and behavioral context over time. The Motional Digital Twin is the output delivered to operators: a live 3D replica of physical flows that consumes what the Dynamic World Model produces. The distinction is similar to the difference between a database engine and the application that queries it; one computes, the other presents.

  • Why do mobile robots still have blind spots even with onboard LiDAR?

    Onboard sensors on a mobile robot perceive only what falls within the robot's current field of view at its current position. Occlusions from shelving, columns, other vehicles or crowds hide objects until the robot is close enough to see around the obstacle. Because the perception window moves with the robot, there is no persistent memory of what was behind an obstacle a few seconds ago. Infrastructure-mounted sensors covering the whole site do not have this problem: they track every entity continuously from fixed vantage points, and that site-wide picture can be shared with any robot on the network in real time.

  • Can a CPU-based edge AI deployment scale across hundreds of sites without a GPU in every location?

    CPU-based inference on current Xeon-class processors is sufficient for the matrix operations required by 3D LiDAR perception when hardware acceleration blocks like Intel AMX are present. The practical implication is that a Physical AI deployment does not require a discrete GPU at each location, which meaningfully lowers per-site hardware cost, power draw and thermal management requirements. For operators managing dozens or hundreds of distributed sites, such as retail chains, hospital networks or multi-airport groups, the total cost of ownership difference between GPU-per-site and CPU-per-site architectures is substantial.

  • What data sources beyond LiDAR can a Physical AI platform fuse at the edge?

    Infrastructure-based Physical AI platforms designed for multi-modal fusion can ingest camera feeds, Wi-Fi positioning signals, boarding pass or ticketing data, point-of-sale records and operational databases such as airport AODB systems. Each additional source attaches attributes to the anonymous spatial ID already established by LiDAR tracking: a camera feed can add an appearance attribute, a boarding pass scan can add a flight association, a POS event can add a dwell-and-purchase link. The LiDAR backbone provides the persistent spatial ID that ties these heterogeneous sources together into a single operational reference.