Skip to content

Total Learning Architecture (TLA)

Data Pillars and Their Applicability to Adaptive Instructional Systems

Overview

The Advanced Distributed Learning (ADL) Initiative has developed the Total Learning Architecture (TLA) since 2016 as a comprehensive 4-pillar data strategy for managing lifelong learning across organizations. The TLA provides a foundational framework that enables interoperability, data-driven decision making, and adaptive instructional capabilities through standardized data management and sharing.

The Four TLA Data Pillars

The TLA is built upon four core data pillars, each based on IEEE standards:

1. Experience API (IEEE P9274.1 xAPI)

  • Tracks and manages learner performance both inside and outside learning activities
  • Uses Learning Record Stores (LRS) to capture learning activity streams
  • Includes xAPI profiles like cmi5 and the TLA Master Object Model (MOM)
  • Enables detailed tracking of learner interactions and performance data

2. Learning Activity Metadata (IEEE P2881)

  • Describes learning activities and their associated content
  • Stored in the TLA's Experience Index (XI)
  • Builds upon IEEE 1484.12.1 Learning Object Metadata (LOM)
  • Increases granularity of learning resource definitions
  • Harmonizes with multiple educational data standards

3. Reusable Competency Definitions (IEEE 1484.20.1 RCD)

  • Describes Knowledge, Skills, Abilities, and Other behaviors (KSAOs) required in the workplace
  • Enables common language for describing competencies across organizations
  • Defines assessment and evaluation criteria for measuring proficiency
  • Supports alignment between education/training and operational performance

4. Enterprise Learner Records (IEEE Study Group)

  • Tracks and manages each learner's competency levels within the organization
  • Built around a comprehensive data model for learner records
  • Supports AI/ML solutions for intelligent tutoring and learner optimization
  • Maintains evidentiary requirements for data ownership and stewardship

TLA Reference Implementation

The TLA adopts a core/edge paradigm that separates:

Core Systems: - Learning Event Management - Activity and Resource Management - Competency Management - Learner Management - Backend services for distributed operations

Edge Systems: - Learning Management Systems (LMS) - Mobile devices, simulators, intelligent tutors - Any learning delivery technology - Adaptive algorithms and visualization tools

TLA Master Object Model (MOM)

The TLA MOM normalizes data across different learning activities through three key states:

  • Learning Activity State: Tracks learner interactions from initialization to completion
  • Learning Event State: Captures context before/after learning activities
  • Career State: Manages progression through career milestones

Five Control Loops for Adaptive Systems

The TLA enables adaptation at multiple levels through five control loops:

  1. Control Loop 1: Optimize learning within current activity (intelligent tutoring)
  2. Control Loop 2: Optimize progress across multiple activities toward credentials
  3. Control Loop 3: Prioritize credentials/activities for potential jobs
  4. Control Loop 4: Support career field management and trajectory planning
  5. Control Loop 5: Enable establishment of new career paths

Key Benefits

  • Interoperability: Standardized data formats enable seamless integration across systems
  • Lifelong Learning: Comprehensive tracking from individual activities to entire careers
  • Adaptive Capabilities: Data-driven personalization and optimization
  • Evidence-Based: Maintains chain of evidence for competency assertions
  • Scalability: Supports everything from individual learner dashboards to organizational analytics

Key Diagrams for Explaining TLA Architecture

Based on the document content, here are the most important diagrams to illustrate the TLA:

1. TLA Core/Edge Systems Architecture (Figure 1)

  • Shows the relationship between core services and edge systems
  • Illustrates the four data pillars integration
  • Demonstrates external interfaces and data flow

2. TLA Master Object Model Verbs (Figure 2)

  • Visualizes the learning event lifecycle
  • Shows the three states: Learning Activity, Learning Event, and Career
  • Demonstrates data normalization across different systems

3. Learning Activity Metadata Structure (Figure 3)

  • Details the P2881 standard data model components
  • Shows relationships between metadata elements
  • Illustrates course catalog federation approach

4. Reusable Competency Definitions Framework (Figure 4)

  • Maps competency relationships and frameworks
  • Shows connection between competencies, credentials, and learning resources
  • Demonstrates the competency development process

5. TLA Control Loops (Figure 7)

  • Illustrates the five levels of adaptive capability
  • Shows time horizons and adaptation scope
  • Connects initial conditions to mission effectiveness

6. Competency Management Data Flow (Figure 5)

  • Shows how learner data flows from activities through LRS systems
  • Demonstrates competency assertion generation
  • Illustrates integration between all four data pillars

7. Enterprise Learner Record Model (Figure 6)

  • Details the comprehensive learner profile structure
  • Shows integration of personal, employment, credential, and competency data
  • Demonstrates longitudinal learning tracking

These diagrams effectively communicate the TLA's comprehensive approach to data-driven learning ecosystems and would be essential for explaining the architecture to stakeholders, developers, and educators.