Multi-Device Environment Reconstruction for Augmented Reality Collaboration
The aim of this project was to scan an environment for augmented reality collaboration and address the challenge of achieving seamless interaction across multiple devices.
3 Months


The project aimed to enable multiple AR devices to scan and reconstruct a shared environment, facilitating collaborative interactions in augmented reality. The system was designed to capture data from the devices, process it to create a unified 3D point cloud of the environment and enable real-time interaction between users within the reconstructed space.
Enable multiple AR devices to scan and reconstruct a shared environment, facilitating collaborative interactions in augmented reality.


One of the biggest challenges was ensuring real-time performance given the limited capabilities of mobile devices. Additionally, achieving accurate environment reconstruction while maintaining low latency for collaborative interactions presented a significant challenge. This was achieved using a server-client architecture, whereby the task of environment reconstruction was offloaded to a powerful server while the AR devices focused on rendering and user interactions.
While the reconstruction produced highly detailed environment models, the HoloLens struggled to render them in real time, resulting in a suboptimal user experience. Future work should focus on optimising the rendering pipeline and exploring the use of more powerful AR devices, in order to realise the full potential of multi-device environment reconstruction for augmented reality collaboration.