Background

In 2020, Tomsk State University (TSU) moved to hybrid learning by combining onsite and online lectures and seminars. Around 15,000 students and 1,500 professors worked concurrently across online and offline environments.

To set up the educational process, TSU engaged its tech partner in developing the Aktru platform, a Yandex Cloud-based solution that enables recording lectures with live streaming from different cameras focused on the board and the teacher’s presentation. This approach has already brought 175 courses into the new learning format, with 611 lectures uploaded in the last semester alone, totaling over 1 TB of data.

TSU searching for a comprehensive solution

When the pandemic hit in 2020, schools and universities had to move students to online learning. Tomsk State University started searching for a comprehensive solution to set up a hybrid learning format, where some students could attend classes in person and others, connect remotely. The university aimed to give all students live access to classes, while also recording high-quality video lectures and storing those records for long-term use. To address this challenge, TSU reached out to partner companies from the University Consortium of Big Data Researchers. For project implementation, the university engaged a company that specializes in installing and setting up multimedia systems for conference halls, meeting rooms, classrooms, and auditoriums.

Since no solution on the market met all the university’s requirements, the tech partner launched a project to develop a hybrid learning system within the Consortium. This project quickly grew into an independent company named Aktru.

Cloud development

Since TSU needed to provide students with continuous access to all video recordings of the university lectures, they had to arrange storage for large amounts of data while ensuring high access performance. Using an S3-compatible cloud storage was a logical choice.

After reviewing the options available in the market, the company opted for Yandex Cloud. In particular, its Yandex Object Storage enables storing data as per the Federal Law 152-FZ regulations, provides automatic scaling as needed, and maintains consistent write and read times for video lectures without slowing down as the amount of stored data grows.

To support the hybrid learning system, the university also wanted to use Yandex SpeechKit. With the speech recognition technology this tool offered, the students would be able to search for information across video lectures by the keywords from the professor’s speech.

Integrated system platform and cloud services

Aktru is an integrated system platform that supports live lecture streaming, recording, cataloging, and viewing. It is designed for hybrid learning, where students and teachers can attend lectures either from home or from the classroom.

Aktru consists of three main components:

  • Classroom equipment
  • Software
  • Cloud infrastructure in Yandex Cloud

The Aktru’s server infrastructure runs on virtual machines in Yandex Compute Cloud, while Yandex Object Storage handles the storage of TSU video materials. Both the virtual machines and the objects in Object Storage reside across multiple distributed availability zones, thus ensuring high data security. Once video materials are uploaded to the cloud, the system immediately integrates them into the Moodle LMS. Dedicated VM servers handle concurrent transcoding of multiple videos.

Yandex SpeechKit enabled the university to set up an automatic speech recognition system for teachers that provides subtitles in Russian and other languages. This allowed students to search for lectures on the Tomsk State University portal by keywords and start watching from the exact part they needed. This way, students can find the information they search for within seconds, while the built-in video editor allows teachers to edit specific parts of their lectures.

Full automation

Today, TSU is equipped to run a fully automated hybrid learning process, covering both online and offline formats. In 2020, around 15,000 students and 1,500 professors worked concurrently across online and offline environments. Based on the comprehensive Aktru solution and Yandex Cloud services, the system synchronizes with the university curriculum and supports lecture recording with simultaneous streaming from different cameras showing the board and the presentation, if launched by the teacher. This approach has already brought 175 courses into the new learning format, with 611 lectures uploaded in the last semester alone, totaling over 1 TB of data.

Together with Yandex Cloud, we are not just delivering a tech project but changing the way education works. Cloud services enhanced the potential of our solution and made cutting-edge tools more accessible to customers in the education sector.

Vladimir Kolomensky,
Aktru Development Team Leader

So far, 45 classrooms at TSU are ready for holding hybrid-format lectures. The next step is to equip the remaining ones. Alongside this, the Aktru team is building a software-based recorder to install on classroom computers. This will remove the need for installing a standalone recorder and management system, helping the university avoid large-scale equipment purchases. This solution will allow TSU to fully transform its educational process within the next few years.

Supported by tech partners, TSU became the first in Russia to create a new learning environment. This initiative became a cornerstone for further development of the nationwide education system, as many universities are now implementing similar projects.

Opinion

Vyacheslav Goyko,
Head of the Center for Applied Big Data Analysis at Tomsk State University
Vyacheslav Goyko,
Head of the Center for Applied Big Data Analysis at Tomsk State University

The higher education system has to adapt to a new generation of students raised in the digital world. In this context, hybrid learning is a must. The tech approach we selected has already brought 175 courses into the new learning format, with 611 lectures uploaded in the last semester alone, totaling over 1 TB of data. Alongside TSU, 29 other universities from the University Consortium of Big Data Researchers are adopting the same solution.