0
Analyses
—
MCDM Score
0
Surveys
A² Learning Studio v2.0
Progress
0%
⚡
EVNN-DC Decision Lab
AI-Powered Multi-Criteria Decision Analysis for Data Centre Lifecycle Investment Under Uncertainty
A² Learning Studio · Lead Developer: Dr. Arezou Shafaghat (Georgia Tech & KSU) · Co-Developers: Dr. Ali Keyvanfar & Dr. Da Hu (KSU)
What is this platform? A browser-based learning and research tool that teaches you how to evaluate data centre investments using Multi-Criteria Decision-Making (MCDM) under uncertainty. You will learn about risk, sustainability, cost, and resilience — then apply that knowledge to real-world case studies and construction scenarios.
Your learning pathway — 6 steps
1
Pre-Module · Learn the Foundations
Explore 6 data centre lifecycle phases, rank decision criteria by importance, and test your knowledge with an 8-question quiz. ~15 min
2
Case Studies · Analyse Real Projects
Run a full MCDM analysis on real data centre projects (Google, AWS, Apple). See how EVT tail-risk, neural network viability scoring, and TOPSIS ranking work together. ~10 min
3
Construction Scenarios · Make Decisions
Face 5 real construction risk scenarios — procurement crises, supply chain disruptions, schedule compression, extreme heat waves, and sustainability trade-offs. Choose your response and see the MCDM impact. ~15 min
4
MCDM Dashboard & AI Guidance · Deep Dive
Explore your analysis results with radar charts, Monte Carlo simulations, and Bayesian AI recommendations. Test "what if" scenarios like extreme heat waves or supply chain shocks. ~10 min
5
Learner Review · Share Your Feedback
Complete a short 8-question survey and help improve the platform. Your responses are anonymized and help validate the MCDM framework. ~5 min
6
Certificate & Export · Claim Your Results
Download your certificate, export decision traces as CSV/JSON for research, and review all your analyses and scenario decisions. ~5 min
🔑 Key terms you'll encounter
MCDM — Multi-Criteria Decision-Making. A framework for evaluating options across multiple dimensions (cost, risk, sustainability, etc.) simultaneously.
EVT — Extreme Value Theory. Statistical methods for estimating the probability and impact of rare, catastrophic events (tail risks).
TOPSIS — A ranking method that scores options by their distance to the ideal solution vs. the worst solution.
AHP — Analytic Hierarchy Process. A method for assigning importance weights to decision criteria.
Monte Carlo — Running thousands of simulations with random inputs to estimate the range of possible outcomes.
Total estimated time: ~60 minutes · All progress is saved in your browser session
🏛️ Kennesaw State University 🏛️ Georgia Institute of Technology 📊 ISO 31000 · PMBOK 🎓 A² Learning Studio
① Pre-Module
3 activities · ~15 min · Complete each to unlock the next
Progress
0%
Quiz
—
Ranked
No
1
Explore Data Centre Lifecycle Phases
Current ▸
Click each phase to learn about the key decisions, risks, and criteria at that stage. Explore at least 3 of 6 phases to unlock the next activity.
Phases explored: 0 / 6 (need 3 to continue)
Site & Planning
Procurement
Construction
Commissioning
Operations
Retrofit / EoL
2
Criteria Framework & Priority Ranking
Locked ▸
First, review the 30+ decision criteria. Then drag to rank the 6 categories by importance.
Your Priority Ranking
    3
    Knowledge Assessment Quiz (8 Questions)
    Locked ▸
    Select an answer to see instant feedback. Questions advance automatically.
    1 / 8
    Complete all 3 activities to unlock →
    Pre-Loaded Case Studies — EVNN + MCDM Analysis
    ② Case Studies — Select a project card below → click "▶ Run Analysis" → explore charts, scores, and budget breakdown. ~10 min
    After running an analysis, face real construction decisions →
    ③ Construction Scenarios
    5 scenarios · Make decisions one at a time · Need 2+ for certificate
    Decided
    0 / 5
    Cumulative ΔCost
    $0M
    Scenario 1 of 5
    Complete scenarios using the stepper above. After the last scenario, you'll continue to the Dashboard.
    MCDM Dashboard — Multi-Criteria Decision Matrix Under Uncertainty
    ④a MCDM Dashboard — Requires a completed analysis. Explore radar charts, run Monte Carlo simulations, and review TOPSIS rankings for your most recent project. ~10 min
    This dashboard implements AHP-inspired weighted scoring with Monte Carlo uncertainty simulation — generating decision traces for advanced data-driven infrastructure decision methods.

    Weighted Criteria Spider

    Criteria Score vs. Weight Scatter

    Monte Carlo Uncertainty Simulation
    Criteria Weight Configuration (AHP-Inspired)
    TOPSIS Ranking — Last Analysis
    Test how new evidence changes your recommendation →
    AI Guidance System — Simulated Explainable AI with Bayesian Updates
    ④b AI Guidance — Requires a completed analysis. Select an analysis → choose an evidence event (heat wave, supply shock, etc.) → click "Generate Guidance" to see Bayesian updates and recommendations.
    This module simulates an explainable AI guidance system for AI/ML decision traces in infrastructure investment. The Bayesian update mechanism shows how posterior probabilities shift as new evidence is integrated.
    Decisions Traced
    0
    Learner trace records
    Bayesian Updates
    0
    Posterior revisions
    Guidance Quality
    —
    Explanation fidelity
    Evidence-Based Recommendation Engine
    Decision Trace Log
    Share your feedback on the platform →
    New Project Analysis — Full Lifecycle Input
    + New Project — Fill in site, energy, and cost parameters → run a full EVNN-MCDM analysis with custom inputs. Compare against case studies.
    Complete all sections for a comprehensive EVNN + MCDM lifecycle assessment. Inputs feed the EVT tail-risk model, neural network viability estimator, and MCDM scoring matrix.
    A. Site & Capacity
    B. Energy & Sustainability
    530
    0100
    70100
    15
    C. Cost & Risk
    110
    15
    After your analysis, explore construction risk scenarios →
    Learner Review & Feedback
    ⑤ Review — 16 questions across 4 sections. Your responses are anonymized and used for platform validation research. Required for certificate. ~8 min
    Section A — Demographics & Background
    1. Primary professional role
    2. Years of experience in infrastructure or related field
    3. Prior experience with MCDM or decision-support tools
    Section B — Framework & Methodology Validation
    4. How useful was the 6-category MCDM framework for evaluating data centre investment viability?
    5. How well did the 30+ decision criteria cover a real-world data centre evaluation context?
    6. How effective was the EVT tail-risk analysis in revealing risks not visible through standard assessment?
    7. How did the uncertainty modelling (EVT/Monte Carlo) change your site recommendation?
    8. How valuable were the construction risk scenarios for understanding real-world decision trade-offs?
    Section C — Usability & Learning Effectiveness
    9. How intuitive was the platform interface and navigation flow?
    10. Rate the quality of the AI-generated feedback and Bayesian guidance explanations.
    11. How much did your understanding of MCDM under uncertainty improve after using this platform?
    12. What was the most valuable learning moment or feature? (Open)
    Section D — Impact, Adoption & Scalability
    13. How critical are MCDM decision-support tools for managing AI infrastructure investment uncertainty?
    14. Would this platform assist your organisation's data centre planning or educational programmes?
    15. How does this platform compare to other decision-support tools or methods you have used?
    16. What criteria, features, or improvements would make this platform more valuable for your work? (Open)
    Export your analyses and decision traces →
    Reports & Data Exports
    ⑥a Reports — Download case reports, survey summaries, decision traces (CSV/JSON), and MCDM matrices. More analyses = richer exports.
    🔬 Research Value

    Decision-trace datasets advance MCDM frameworks for AI infrastructure decisions. Exportable as structured JSON/CSV for research and publication.

    🌍 Educational Impact

    Aggregated learner survey data, role diversity metrics, and educational module completion rates demonstrate the platform's educational effectiveness.

    📦
    Full Session JSON — Comprehensive POC Data Package
    One file containing ALL data from your entire session: demographics, quiz, rankings, analyses, scenarios, AI traces, survey responses, timing analytics, and interaction log.
    📊

    Case Study Report

    Full EVNN-MCDM validation including EVT parameters, n* configuration, budget, sustainability metrics, and viability narrative.

    📋

    Sponsor Survey Report

    Aggregated survey statistics: market validation, adoption intent, criticality scores, and feature requests.

    🔬

    Decision Trace CSV

    Learner decision traces with criteria weights, uncertainty thresholds, MCDM scores, and scenario outcomes.

    ⚡

    MCDM Criteria Matrix CSV

    Sub-criteria weights, normalised scores, TOPSIS rankings, and Monte Carlo uncertainty ranges per project.

    🤖

    AI Guidance JSON

    Bayesian update traces, evidence events, posterior probability shifts, and recommendation rationale for research validation.

    🏗️

    Construction Scenario Log

    Learner decisions across procurement, scheduling, and resilience scenarios with MCDM impact scores and feedback traces.

    All Analyses This Session
    Generate your completion certificate →
    Completion Certificate
    ⑥b Certificate — Complete all 6 requirements below → enter your name → generate your personalized completion certificate.
    Complete the pre-module quiz, at least one case study analysis, two construction scenarios, and the learner review to unlock your certificate.
    Export your data or return to the beginning