Scaling industrial AR: How Visometry uses Unity and model tracking on the factory floor

From precision tracking to quality inspection, Visometry and Unity are enabling next-gen augmented reality across manufacturing and engineering.Visometry GmbH is a deep-tech company based in Germany, specializing in enterprise-grade augmented reality (AR) solutions for industrial use. A spin-off from the renowned Fraunhofer Institute, Visometry is best known for its powerful model tracking SDK, VisionLib, which enables precise, real-time object tracking in complex industrial environments. From automotive to mechanical engineering, Visometry’s tools help companies deploy scalable AR applications that drive efficiency, quality, and innovation.In this article, Visometry explores how they are leveraging Unity’s real-time 3D engine to bring advanced tracking and 3D visualization to life — from inspection systems like Twyn to immersive interactive experiences, and enabling Visometry to streamline industrial AR development and make high-performance solutions accessible to developers, integrators, and manufacturers alike. Industrial AR today: From pilot to productionAR has quietly transformed the industrial world. Over the past decade, industrial AR has progressed from experimental trials to mission-critical tools that are reshaping how companies inspect, maintain, and interact with complex machinery and infrastructure.As the sector confronts new challenges—rising demand, an aging workforce, and the constant push for precision—AR offers compelling answers to improve operational efficiency, reduce errors, and make smarter use of skilled labor.Visometry has been part of this evolution from the beginning, leveraging Unity’s real-time 3D engine to bring advanced tracking and 3D visualization to life. Originally spun out of the Fraunhofer Institute, the company today focuses on a critical layer of industrial AR: enterprise-ready object tracking. This core technology enables the precise alignment of 3D content and data layers with real-world machines and components. Enabling precise AR with VisionLibAt the heart of many AR solutions is VisionLib, a software development kit (SDK), built on Unity, that brings industrial-strength object tracking to AR applications. Already in use by leading companies across automotive and mechanical engineering sectors, VisionLib provides stable, real-time, markerless tracking of one or more physical objects—even under challenging lighting or motion conditions.Using CAD and 3D data, VisionLib enables high-precision pose estimation that anchors digital content directly onto real-world components. This technique, known as model tracking, has become a foundational capability for industrial AR, enabling developers to overlay visual data with accuracy and without manual alignment or markers. Unlike consumer-grade AR, which typically relies on simultaneous localization and mapping (SLAM) -based camera tracking and can suffer from content drift, model tracking ensures precise placement of 3D elements even in dynamic or reflective environments. That level of precision is key for industrial use cases such as quality inspection, guided assembly, and training.From SDK to solution: Twyn for AR-based quality inspectionWhile VisionLib is designed for developers and platform integrators, Visometry also brings AR directly to manufacturing teams through Twyn — a ready-to-use AR inspection solution.Twyn enables interactive, CAD-based quality inspection on the shop floor, allowing teams to detect deviations between digital models and physical parts quickly and flexibly. By combining VisionLib’s tracking with Unity’s visualization engine, Twyn delivers enterprise-ready performance in an intuitive interface, even for teams without deep expertise in 3D graphics or AR development.This approach exemplifies Visometry’s mission: democratize access to industrial-grade AR by abstracting complexity and enabling other organizations to offer offering scalable solutions that work out of the box. Why Unity: Developer speed and cross-platform reachFrom the outset, VisionLib has been designed with developers in mind. Leveraging Unity's plugin architecture, VisionLib SDK seamlessly integrates into the Unity ecosystem, enabling developers to create industrial-grade AR applications with precise object tracking and robust anchors for digital content, all managed directly from the Unity tech stack.Unity has long played a pivotal role in democratizing model tracking, particularly through its ARFoundation framework, which bridges platform-specific AR modules like ARKit and ARCore.Building on this foundation, VisionLib SDK extends Unity's capabilities to meet the demands of industrial applications. Designed for enterprise-grade augmented reality, VisionLib combines CAD data with advanced image processing to deliver true 3D object tracking for mixed and augmented reality scenarios.Unity developers can harness this cutting-edge technology directly within the Unity Editor to craft high-quality XR experiences. Moreover, VisionLib SDK integrates seamlessly with ARFoundation, enhancing interoperability and expanding its reach. This streamlined workflow empowers developers to rapidly prototype, efficiently implement AR features, and deploy across a wide range of devices, including smartphones, tablets, AR glasses, and stationary industrial hardware. Real-world AR: A closer look at Atelier MarkgraphOne standout industrial use case is Atelier Markgraph, a design studio specializing in spatial communication. For years, they’ve been using AR to stage products and complex systems in dynamic, interactive ways—often at high-profile events.Working with Mercedes-Benz, Markgraph created AR-driven experiences to showcase vehicles and engines. Rather than display a static engine block, they used VisionLib and Unity to deliver layered, explorable, and interactive 3D visualizations.“It’s super complex to physically cut through an engine,” explains Christoph Diederichs, Head of Interactive Experiences at Markgraph. And it’s difficult to convey engineering finesse just by presenting a block of steel – it’s not interactive, it’s not dynamic.”According to Diederichs, AR makes it possible to add an interactive digital 3D layer to the real world, enabling spectators “to grasp” things that are otherwise hard to communicate in a short amount of time. Tracking and interacting with a real object delivers on the “tangible feel” and the appeal of a real engine, while AR is virtually making the invisible visible. “In our industry, time is always a critical factor,” says Diederichs. Especially during the final stages of a project, when everything needs to come together under pressure: 3D models need to be swapped, last-minute change requests incorporated."At one event, there was virtually no time for testing: …and it just had to work”, Diederichs recalls. The new engine, still a closely guarded secret until its public unveiling, could only be viewed and tested once in advance. The rest had to be developed under lab conditions only. The future of industrial AR: From guidance to validationAccording to Jens Keil, founder and product manager at Visometry, AR in industrial settings is entering a new phase. The early days were about enabling tracking. Then came platforms for scaling content. Now, there is a shift toward using AR not just for guidance, but for verification.That means going beyond simply showing what to do—and instead confirming whether a task was performed correctly. VisionLib is already ahead of the curve with multi-model tracking, which can independently track and validate multiple objects or subcomponents within a larger assembly.In one example, Visometry highlights how its tracking tech automatically detects and flags misaligned components on steel structures. Combined with advances in AI and machine vision, this approach moves AR toward more intelligent, closed-loop validation tools.“We’re exploring ways to enable users to combine AR, AI, and other techniques,” says Keil. “Our goal is to democratize access through simple developer tools and platform-independent solutions.”Visometry is not just creating tools for today’s AR but building a foundation for the next generation of spatial computing—where AR, AI, and computer vision come together to empower industry-wide transformation.