Final Infographic – Project Summary & Insights

At the end of the three challenges, students are asked to produce a single-page
infographic (digital poster ) that summarises the overall project.

The infographic should be understandable to a non-technical audience (e.g. city
officials or citizens) while still being grounded in the analyses performed.

6.1 Required Content of the Infographic

  • Title and short context: a concise title and 2–3 sentences explaining the smart-city energy & fairness scenario.
  • Data sources: mention the smart-meter energy dataset (OPSD) and the socio-economic fairness dataset (Adult), and briefly describe their roles.
  • Pipeline overview: a simple visual flow or diagram showing the three main phases:
    • – Challenge 1: energy data cleaning, daily aggregation, and outlier detection.
    • – Challenge 2: fairness evaluation and fair sampling using AIF360 and SQL.
    • – Challenge 3: regression modelling and fairness of prediction errors.
  • Key quantitative results:
    • – Example outlier pattern from Challenge 1 (e.g. unusually high consumption days).
    • – At least one fairness metric from Challenge 2 (e.g. statistical parity difference or disparate impact before/after Reweighing or SQL sampling).
    • – At least one modelling metric from Challenge 3 (e.g. RMSE and R² for both models).
  • Fairness-of-errors summary: a simple table or graphic comparing group-wise MAE across building_type and/or location, and (optionally) the fairness gap/ratio for each model.
  • Main insights and recommendations:
    • 3–5 bullet points interpreting what the city should learn from the analysis (e.g. which buildings are most demanding, where errors are larger, how fairness issues might appear).
    • At least one suggestion for improving data collection, modelling, or policy design to support more equitable and efficient energy management.

Infographic Deliverable

Students may use any tool (PowerPoint, Keynote, Canva, Figma, etc.) to create the infographic, but it should:

  • Fit on a single page or slide (A0 or similar).
  • Be visually clear and readable when printed or projected.
  • Include legends and units for all plots or numbers.
  • Explicitly connect energy analytics and fairness considerations.