KA Lite to Kolibri

New infrastructure for offline learning

L E A R N I N G

E Q U A L I T Y

Kolibri is an open-source ed-tech platform designed for low-resource communities

Repository of open content, with tools to support curation

Seamless distribution of content and aggregation of usage data

Runs on a wide variety of hardware platforms

• Only Khan Academy content, aligned to US standards �
• Not easy to integrate new content sources�
• No way for end users to adapt and align curricula to match local needs and standards�

​

• Ecosystem of public channels, aligned to multiple standards�
• New content can be directly uploaded or imported via API�
• Web-based alignment tools enable curators to combine and rearrange content from multiple channels�

→

Kolibri consists of two software components

The Kolibri application is installed onto a local device, and imports content channel.

The Content Curation Server is a central cloud server that aggregates content from many sources, organized into content channels.

NGOs, MOEs,

Authors, Teachers,

content experts

Teachers and

Students

Teachers and Administrators

Content channels

A channel is a collection of content and metadata, organized into a

tree structure

​

Other channels are manually curated through the web interface

Content relationships:

- related

- prerequisites

- tagged with labels

Channels can be either

- unlisted

or

- public

Channels can be programmatically created

Folder nodes:� - TopicNode

Content nodes:� - AudioNode� - DocumentNode� - HTML5AppNode� - VideoNode� - ExerciseNode

Content Curation Server

Main channels

Specialized channels

(partners create & maintain custom forks)

Kolibri Content Pipeline

Sushi Chefs

scraping scripts

and channel upload

ricecooker

python utils for uploading channels

API

Broad hardware and OS support

Kolibri runs on a wide variety of hardware and software platforms

… or on a device acting as a local server, accessible by browsers on nearby client devices, either via a hotspot connection or a Local Area Network in a computer lab.

Kolibri can run as an app directly on the end-user’s device, for standalone, offline usage, also able to sync and share with other nearby devices that are running Kolibri...

Content distribution and data aggregation

Designed to overcome the challenges faced in low-bandwidth and offline contexts,�Kolibri facilitates seamless content distribution and aggregation of usage data.

Nearby devices running Kolibri can connect peer-to-peer, over Wi-Fi or other local network, sharing content or syncing usage data.

Content and usage data can be exported from Kolibri onto portable USB storage media, and then carried by foot and imported to other devices.

Kolibri can opportunistically leverage any Internet connectivity that exists, to bring in updates and sync data back to a central server.

content distribution

data aggregation

Timeline

Currently

• Active software development (10x devs)
• Content efforts to create channels (4x content-devs)
• Implementations and teacher training (6x implementations)

​

Plan

• Deployed and tested by partners early Fall 2017
• Public launch in late Fall 2017
• DEMO: https://kolibridemo.learningequality.org

​

​

​

• KA Lite project continues (bug fixes but no new features)
• Organizations interested in multi-channel support and tools for curation should look at Kolibri and Content Curation Server

​

Guillaume Duchayne

Hea Poh Lin

Icon Fair

Krisada

Marie Van den Broeck

Oliviu Stoian

Pham Thi Dieu Linh

Pravin Unagar

Ralf Schmitzer

Travis Avery

Unlimicon

Yu Luck

Arthur Shlain

Aybige Aya

Carlos Salgado

Chameleon Design

Claire Jones

Creative Stall

Cristiano Zoucas

Davo Sime

Ed Scotti

Futishia

Gan Khoon Lay

Gregor Črešnar

Image Credits

https://thenounproject.com/

learningequality.org/kolibri

learningequality.org

We're ready to fly!

github.com/learningequality