What is CARL

CARL is a Unity-Application for the Microsoft HoloLens, developed by Asmus Eilks and Dr. Felix Putze. It allows users to enter a shared virtual space together - even if they are in entirely different real spaces.
This can be used to accomplish almost any design or coordination task, and the integrated data-collection to LabStreamingLayer allows for the collection of high fidelity motion data about the activity.

Key Features

Intuitive

Controller-Free interaction through Handtracking via MRTK requires almost no initial tutorial

Mobile

Spatial synchronization is done via QR-Codes, which can be printed out on a piece of paper and placed anywhere

Extensible

The application was developed with extensibility as a key goal from the start, making integration of new features as easy as possible

Integrated Data collection

Hand & Headtracking as well as Events are automatically logged to the LabStreamingLayer

Detailed Interaction

Unlike VR-Controllers, Hand-Tracking and real, motion-tracked objects yield accurate data with high fidelity.

MIT-License

Use anywhere, for anything.

A small validation study (n=12, split in 6 groups of 2), was conducted to test an aspect of the systems design. While the study showed no significant difference for the investigated interaction element, the application scored an average of 5.06/7 on the SWUS-Scale, which assesses the usability of shared workspace groupware applications.

Works with HoloLens 2 & Optitrack Motive, as well as a Unity-Editor based server on any windows machine.

Repository Size: Compressed ~600mb, after Unity-build/precompilation about 6GB

Rendering performance dependent on HoloLens Device & Unity, based on a test by Fologram up to 2 million polygons are feasible.

Tested with an update rate of 30Hz, synchronization latency dependent on network.

~3kB/s bandwidth needed per constantly moving object. Even a slow connection of ~500kB/s could therefore support dozens of moving objects, and keep them synchronized.

A detailed developers-manual can be found here.
The application itself can be cloned from this github-repository.
The original thesis that CARL was developed for can be downloaded here.