Building Multi-User AR on iOS with Collaborative Session

A client asked us to develop an AR app for collaborative interior viewing: two designers with iPads simultaneously placing furniture in a shared virtual space. Without synchronization, each saw a different set of objects, and the real room ended up in different coordinates. The solution — Collaborat

Development and support of all types of mobile applications:

Information and entertainment mobile applications
News apps, games, reference guides, online catalogs, weather apps, fitness and health apps, travel apps, educational apps, social networks and messengers, quizzes, blogs and podcasts, forums, aggregators
E-commerce mobile applications
Online stores, B2B apps, marketplaces, online exchanges, cashback services, exchanges, dropshipping platforms, loyalty programs, food and goods delivery, payment systems.
Business process management mobile applications
CRM systems, ERP systems, project management, sales team tools, financial management, production management, logistics and delivery management, HR management, data monitoring systems
Electronic services mobile applications
Classified ads platforms, online schools, online cinemas, electronic service platforms, cashback platforms, video hosting, thematic portals, online booking and scheduling platforms, online trading platforms

These are just some of the types of mobile applications we work with, and each of them may have its own specific features and functionality, tailored to the specific needs and goals of the client.

Showing 1 of 1All 1734 services
Building Multi-User AR on iOS with Collaborative Session
Complex
~1-2 weeks

Our competencies:

Frequently Asked Questions

Latest works

  • image_mobile-applications_feedme_467_0.webp
    Development of a mobile application for FEEDME
    895
  • image_mobile-applications_xoomer_471_0.webp
    Development of a mobile application for XOOMER
    782
  • image_mobile-applications_rhl_428_0.webp
    Development of a mobile application for RHL
    1216
  • image_mobile-applications_zippy_411_0.webp
    Development of a mobile application for ZIPPY
    1079
  • image_mobile-applications_affhome_429_0.webp
    Development of a mobile application for Affhome
    1002
  • image_mobile-applications_flavors_409_0.webp
    Development of a mobile application for the FLAVORS company
    597

A client asked us to develop an AR app for collaborative interior viewing: two designers with iPads simultaneously placing furniture in a shared virtual space. Without synchronization, each saw a different set of objects, and the real room ended up in different coordinates. The solution — Collaborative Session from ARKit. But implementation turned out to be nontrivial: MultipeerConnectivity periodically lost connection, maps didn't merge, and objects jumped. Below — how we build a reliable multi-device AR session, what pitfalls we avoid, and what to consider before starting development. Development timelines and pricing are determined individually.

Collaborative Session allows multiple iOS devices to exchange spatial maps in real time. Unlike Cloud Anchors, the internet is not required — data is transmitted over the local network with latency under 20 ms. We implemented a project for a design studio where 3 participants synchronized with latency below 15 ms; server cost savings amounted to around $1500 per month. However, this technology only works on Apple devices and is sensitive to network quality.

How Collaborative Session Works

Activating Collaborative Session requires just one line in the configuration:

let config = ARWorldTrackingConfiguration() config.isCollaborationEnabled = true arView.session.run(config) 

After that, ARKit generates ARSession.CollaborationData packets — information about the device's environment. These need to be sent to all participants via a suitable transport: MultipeerConnectivity, WebSocket, or GameKit.

func session(_ session: ARSession, didOutputCollaborationData data: ARSession.CollaborationData) { let encoded = try? NSKeyedArchiver.archivedData(withRootObject: data, requiringSecureCoding: true) // send encoded to all peers } // On the receiving side: let data = try? NSKeyedUnarchiver.unarchivedObject(ofClass: ARSession.CollaborationData.self, from: received) session.update(with: data) 

After 10–30 seconds, devices merge their maps — all participants get a unified coordinate system.

Why Collaborative Session Is Faster than Cloud Anchors

Collaborative Session is a direct alternative to Cloud Anchors. Here's a comparison:

Parameter Collaborative Session Cloud Anchors (ARCore)
Latency <20 ms (local network) 100–500 ms (via server)
Internet Not required Mandatory
Number of devices Up to 4–5 Virtually unlimited
Platform Only iOS (ARKit) Android, iOS (via ARKit+ARCore)
Server cost None Google Cloud (paid traffic)

For local scenarios (office, exhibition), Collaborative Session is 30–50% faster and requires no external servers. Infrastructure savings reach 50–60%.

How to Ensure Reliable Synchronization

  • MultipeerConnectivity is unreliable. MCSession drops peers without clear errors. Solution: a MultipeerConnectivity heartbeat every 2 seconds, automatic reconnection after three missed pings.
  • CollaborationData packet size. With active camera movement, traffic reaches 50–100 KB/s per device. For 4 participants — up to 400 KB/s. For 5 devices, it can approach 500 KB/s. Send CollaborationData over .unreliable (ARKit recovers from lost packets). Send critical data (object coordinates) over a separate reliable channel.
  • Heartbeat mechanism. Our certified ARKit developers with 7+ years of experience guarantee that MultipeerConnectivity heartbeat ensures stable connections.

What Happens When Maps Merge?

Before merging, each device has its own local coordinate system. Anchors added earlier may jump. Track ARParticipantAnchor:

func session(_ session: ARSession, didAdd anchors: [ARAnchor]) { let participants = anchors.compactMap { $0 as? ARParticipantAnchor } if !participants.isEmpty { // Maps merged — can now place objects enableObjectPlacement() } } 

Place objects only after ARParticipantAnchor appears — otherwise desynchronization is inevitable.

What Data Is Synchronized in Collaborative Session?

CollaborationData contains feature points, depth map, and anchors. Custom objects (e.g., virtual furniture) are not synchronized automatically — their coordinates must be sent separately. We recommend a reliable channel for object positions and .unreliable for CollaborationData.

Data Type Channel Mode
CollaborationData MultipeerConnectivity/WebSocket Unreliable
Object coordinates Separate channel Reliable
Events (placement, deletion) Separate channel Reliable

Our Process: From Analysis to Deployment

  1. Analyze use case and network constraints.
  2. Design architecture: transport selection, sync scheme.
  3. Implement ARWorldTrackingConfiguration and transport layer.
  4. Handle session lifecycle: participant discovery, map merging, peer departure.
  5. Test on real devices under various network conditions.

What's Included in Our Work

  • Collaborative Session architecture with transport selection (MultipeerConnectivity, WebSocket, GameKit).
  • Implementation of heartbeat, reconnection, and buffering.
  • Synchronization of custom objects via a separate reliable channel.
  • Integration with your UI: placement, deletion, movement of objects.
  • Documentation and source code access.
  • Team training (2 hours online) and post-release support (7 days guarantee).

Timelines

A basic Collaborative Session with MultipeerConnectivity — 2–3 weeks. With full edge‑case handling, state synchronization, and UI — 4–7 weeks. Development cost starts at $5,000 for basic integration. Consultation available upon request.

Common Implementation Mistakes

  • Packet loss under unreliable delivery is not critical for CollaborationData — ARKit rebuilds the map.
  • Reliable delivery of CollaborationData causes queuing and lag — use .unreliable.
  • Placing objects before map merge leads to anchor jitter — wait for ARParticipantAnchor.
  • A departing participant without buffer clearing accumulates packets — clear the buffer on disconnect.

We are ready to help with your AR project.