Overview
Stellar is the brand name for MIT's platform for course management and learning. Currently, this platform consists of a home-grown Learning Management System (LMS), Stellar 2.x, which has had an admirable history of intensive usability and accessibility testing and high responsiveness to user feedback. However, the current Stellar LMS is currently at a critical juncture in terms of scalability and applicability to the existing landscape of teaching and learning at MIT. In simple terms, the Stellar platform, launched in 2001, is showing signs of age.
Key issues facing the current platform include:
- Technical and usability issues resulting from organic (rather than planned) and simultaneous growth on many fronts.
- Integration and functionality issues resulting from the inclusion of disparate third-party solutions and their interfaces.
- Code compatibility and stability issues resulting from partial, time-constrained and resource-dependent code updates for selected components.
- Architectural issues resulting from the lack of a unified vision or strategy.
The current Stellar LMS is also limited by major service hindrances on two fronts:
- The current platform's hard-coded architectural limitations and resultant pedagogical shortcomings undercut its core mission as a Learning Management System
- Its operational overhead for support and maintenance exceeds that of other possible alternative platforms
While the latter issue primarily affects development and support staff, it nevertheless feeds into Stellar 2's instructional platform and user support challenges. The constraints imposed by the architecture - coupled with the fragility of the code and the high resource overhead required to alter it - limit the upgradeability of the current LMS while engendering user dissatisfaction with its limited extensibility. Work-arounds designed to circumvent these service gaps necessitate resource-intensive manual intervention by developers. While they do fulfill use cases not accounted for within the application's main feature set, such work-arounds also generate additional overhead. All of these issues must be addressed if Stellar as a product is to come into its own as an Institute-wide Learning Management System worthy of MIT faculty and students. In Spring 2009, IS&T evaluated several products and services – Moodle 1.9, MoodleRooms, Drupal 6, Sakai versions 2x and 3, and BlackBoard versions 8 and 9 – against functional requirements, data dependencies, and other key criteria for an LMS.
Background
Stellar 2 was re-scoped in the winter of 2008 to align with a new strategy of provisioning an industry-standard, enterprise-class, scalable LMS solution capable of supporting teaching and learning at MIT. Operational stability of the current Stellar platform, along with refinement of user-facing tools in a forward-compatible manner, has remained a top priority throughout the process of investigating alternatives to the current back-end architecture. Simultaneously, IS&T undertook an evaluation of LMS solutions upon which to build the next generation of Stellar. The goal: to find the best combination of features and functionality that could be supported with a reasonable commitment of resources.
After verifying and documenting key Stellar use cases and features, IS&T worked with user and stakeholder groups at MIT to validate requirements for a next-generation LMS for the Institute. In Spring 2009, IS&T evaluated several products and services in use at peer institutions against these requirements, as well as functional specifications, data dependencies, and other key criteria specific to MIT. In Summer of 2009, IS&T presented its findings to user and stakeholder groups. The initial LMS evaluation was concluded on June 1st 2009, with the recommendation that MIT pursue one of three possible scenarios: build, wait, or buy.
In Fall 2009, the Faculty Advisory Committee on Learning Management Systems, comprised of faculty, students, and staff, was formed to refine and extend the evaluation of stakeholder needs. This committee is working with IS&T to determine relevant criteria, validate data from inside and outside of MIT, re-examine viable platforms, and provide direction on Stellar NG, the next generation of Stellar at MIT.
Next Steps
MITCET has empowered the Faculty Advisory Committee on Learning Management Systems with a faculty chair and appropriate representation from the various schools at the Institute, with the goal of assisting IS&T in choosing the best option for the MIT community.