AI Avatars & Spatial Computing · Mixed reality
Spatial Computing in the Enterprise: Where It Pays Off Today
Spatial computing connects digital information with physical space: 3D models standing in the room, instructions right at the machine, exercises captured spatially by a camera. After the metaverse hype, it is becoming clear where the technology actually moves companies forward.
Julia Rose
Key takeaways
- Spatial computing combines computer vision, sensors, 3D data and AI with headsets, depth cameras or screens to bring digital content into real space.
- Today it pays off mainly where spatial understanding makes the difference: MedTech, training and service.
- Hardware changes: Microsoft has discontinued HoloLens 2, while Meta Quest and Apple Vision Pro are widespread. Solutions should not be tied to one device.
- At ML Con Munich 2026, our CTO Agata Chudzińska showed why spatial understanding shapes the next stage of AI.
What spatial computing is
Simon Greenwold coined the term in his 2003 work at MIT: interaction between people and machines in three-dimensional space. Today spatial computing covers all technologies that place digital content in space and connect it with the real environment:
| Building block | Role |
|---|---|
| Glasses and headsets | show digital content in space, such as Meta Quest or Apple Vision Pro |
| Depth cameras and sensors | capture space, body and movement in 3D |
| Computer vision | recognises objects, people and poses in the images |
| 3D data and digital twins | represent machines, buildings or anatomy as spatial models |
| AI | understands language, space and processes and controls the interaction |
| Edge computing | processes data on the device, without cloud latency |
Where it pays off today
MedTech: anatomy in space
Surgeons plan procedures with 3D models built from the patient’s MRI and CT data and displayed in space through an XR headset. For the model to exist, the structures have to be segmented first.
In practice: apoQlar
The VSI HoloMedicine® platform shows a patient’s anatomy in 3D through an XR headset, Meta Quest devices among them. Our AI models segment the structures from MRI and CT automatically, in seconds instead of hours, and the result goes straight into the headset.
Training and rehab: practice with spatial feedback
Where movements have to be performed correctly, depth cameras capture the body in 3D and VR elements make practice tangible. That applies to rehab exercises as much as to manual tasks employees learn.
In practice: 3D movement analysis
In a research partnership, several Azure Kinect cameras capture patients doing rehab exercises at home. The system checks the exercise in real time against the physiotherapists’ criteria, and VR game elements help patients keep going.
Service and visitor guidance: information where it is needed
Instructions right at the machine, remote support or orientation at large events: spatial computing brings information to where the work happens. A well-known solution for this was Microsoft Dynamics 365 Guides and Remote Assist on HoloLens 2. Microsoft ended HoloLens 2 production in October 2024, and both applications are no longer available after 31 December 2026. Anyone planning such applications should therefore look for device-independent solutions.
In practice: German Congress of Surgery
At the 141st Congress of the German Society of Surgery, a voice assistant answered questions about the programme and venue for four days, on a screen kiosk and in VR on Meta Quest 3.
Why spatial understanding matters more for AI
At ML Con Munich on 25 June 2026, our CTO Agata Chudzińska gave the keynote “From Language to Reality: Why the Future of AI Is Spatial”. The thesis: scaling language models shows diminishing returns, and the next advances will come from systems that understand space, movement, and cause and effect. For companies this means that wherever AI has to understand physical processes, spaces or bodies, the value of spatial computing keeps growing.
More in the article ML Con Munich: Why the future of AI is spatial.
What adoption requires
- A clear use case: a task where spatial understanding helps measurably, such as planning time, error rate or training time.
- Device independence: build content and logic so that changing headsets does not mean starting over.
- Integration: 3D data, instructions and results come from existing systems and flow back to them.
- Compute in the right place: real-time tasks run on or near the device, heavy analysis in the cloud.
- Data protection: cameras capture people and spaces, which has to be planned from the start.
How AI avatars are used in VR and mixed reality is covered in our article AI avatars in VR and mixed reality.
Spatial computing for your use case?
We look at where spatial data and XR help measurably in your processes and what a device-independent solution looks like.
Request a process analysisFrequently asked questions
Technologies that place digital content in space and connect it with the real environment, such as XR headsets, depth cameras, computer vision, 3D data and AI. The term goes back to Simon Greenwold and his 2003 work at MIT.
Today mainly in MedTech, for example surgical planning with 3D models, in training and rehab with spatial feedback, and in service and visitor guidance.
No. Microsoft ended HoloLens 2 production in October 2024. Dynamics 365 Guides and Remote Assist are no longer available after 31 December 2026.
AI recognises objects, people and movement, understands language and controls the interaction. Researchers also see spatial understanding as the next big step in AI development.
Julia has been part of theBlue.ai since 2019 and has accompanied the development of AI applications in the enterprise environment since the company’s early days. In her role as Marketing Lead, she works closely with the engineering and consulting teams and makes complex technical topics understandable and accessible for decision-makers.
In her articles, she writes about practical experience from enterprise AI projects, as well as the challenges and opportunities of using AI in companies.