Elabor8 worked with The University of Melbourne on the highly successful project to move from the aging Blackboard Learning Management System (LMS) to a new cloud-based system, Canvas. This change was required to allow The University of Melbourne to better support innovation in pedagogy and curriculum, through the ability to create flexible, interactive and engaging learning spaces, and incorporate a growing suite of digital tools.
Canvas LMS was successfully launched as the new Learning Management System across The University of Melbourne for the first semester in 2020. Elabor8 helped The University of Melbourne to embed Atlassian tools and agile practices with the teams on the project, but the focus of this case study is on the somewhat novel use of Jira and Confluence to manage the migration of the 6000 subjects from the incumbent LMS to the new system.
A technical and human challenge
As part of the project, The University of Melbourne developed a sophisticated automated migration tool to copy and translate the subjects from Blackboard to Canvas. But ensuring that subjects were ready for teaching in the new LMS was both a technical and human challenge.
The program needed a way to manage, track and provide visibility over the large number of subjects as they moved through the process of automatic migration, collaboration and consultation with subject coordinators, manual remediation, integration and configuration of subject-specific digital tools, and acceptance/handover to the faculty.
By using Atlassian Jira, and paired with Atlassian’s collaboration tool, Confluence, we were able to effectively manage the migration process, distribute the workload and provide transparency to the team and stakeholders. Below are some of the key implementation details, describing how we used the Atlassian products to our advantage.
Single point of truth
All 6,000+ subjects were created in Jira as a custom issue type (“Subject”) via an imported spreadsheet, with data collated from the University’s subject administration system.
Elabor8 configured Jira (Cloud) to support having each subject represented in Jira, giving the program a self-contained point of truth for the subject as it moved through the migration process, including:
- The status of the subject in the workflow
- Link to the original and migrated subject
- Attached report detailing the auto-migration results
- Link to the Confluence page detailing actions that came out of consultation with the subject coordinators
- Meta-data relating to the subject that allowed for targeted concierge, reporting and troubleshooting (using both standard Jira fields and custom fields)
- Dependencies, such as subject-specific tools to be integrated and configured
Small batches and decentralised management
Using Jira’s “Versions” to indicate ‘transition windows’, subjects were split into large groups based on the teaching periods they were required for. This allowed the migration team to prioritise, track and report on a large number of subjects effectively.
Elabor8 set up Jira with epics to represent faculties, and a Scrum Board per faculty, with a common sprint used in parallel across all 10 faculty boards. Subjects that had been automatically migrated, were allocated to a Sprint by the team lead, and were then picked up by one of the team of PhD students working as Migration Support Assistants (MSA), to perform the consultation and manual remediation work.
The flexibility in using Jira’s Scrum Boards and sprints gave the migration team the ability to create smaller batches, distribute work easily, manage work in progress, and to set tangible and measurable targets.
By using Jira’s workflow functionality the program was able to incorporate the automated and manual tasks into a cohesive process. Automated migration details were updated in Jira via the API, allowing team members to easily locate and pick up subjects that were ready for remediation. The details and status of the subject was managed and updated by the person doing the work.
This approach avoided delays, overheads and bottlenecks that would have occurred if we relied on a centralised role to manage. Jira, and the information in Confluence, facilitated an easy handover between MSAs when needed, and eventual handover to the Faculty.
Real-time reporting and fact-based forecasting
A key success factor for the project was the ability to easily report on the current status of the migration, and to use actual data to make accurate forecasts for the remaining migration work. This allowed the program leadership to make data-driven decisions to give the project the best chance of success.
A script was developed to analyse the subjects in the source LMS system, and automatically attributed story points in jira (via API), based on their complexity 2 (Low), 5 (Medium) or 13 (High). Using Tempo for Jira (an Atlassian Marketplace add-on), the MSA’s logged the time spent on that subject.
Elabor8 analysed the story points and the time spent data – and was soon able to provide a very accurate estimate for how much effort was required for each subject and therefore the entire migration effort (the backlog).
Using this data, Elabor8 was able to provide information about how the migration was tracking and what adjustments could be made to ensure all subjects would be available and student-ready by the required teaching period. Continued monitoring of this data confirmed the approach, and so the program leadership was able to confidently report to the steering committee.
Alignment and transparency
The use and promotion of Jira dashboards – including wallboards, gadgets and filters – was invaluable in communicating the plan and progress to the migration teams, the broader project teams and to stakeholders.
Different dashboards were set up to provide the most relevant real-time information for the leadership team, the leaders of the migration efforts and the Migration Support Assistants.
The use of Sprints gave us a smaller, defined goal to communicate and track, while still ensuring the whole ‘version’ was on target. Confluence was also used to publish daily and end-of-sprint progress updates.
This transparency provided the migration team with a great understanding of the whole of the mission, the size and complexity of the project, while empowering them to take ownership and strive for excellence in their crucial area.
Effective management of the migration process
By using Atlassian Jira and Confluence, Elabor8 rapidly configured a platform that met the specific needs of managing the migration of more than 6000 subjects through a process that incorporated large batch automatic migration, small scale face-to-face consultation and bespoke remediation.
The program was able to easily distribute the workload and provide transparency to the team and stakeholders, ensuring that we could confidently meet ambitious targets and deliver the subjects into the new LMS for the faculties and students of The University of Melbourne.