Challenge 1: City-Business Collaboration

To identify opportunities for companies and cities to collaborate on sustainability solutions, we need to first understand how the data they each report is aligned or divergent by using data science and text analytics techniques.

1. Utilize data science and text analytics techniques to improve readability of the data.

2. Visualize shared sentiment between cities and companies on specific topics (such as clean energy, sustainable buildings, clean transport, waste and circular economy).

3. Create a KPI model that measures the propensity of cities to collaborate with companies. Provide the justification of your weights and indicators.

[See the Hackathon Overview Deck linked above for more information on the challenges.]

Challenge 2: Water Security/Climate Risk

To help companies and cities fully understand their current water security and future climate risks, we must first inspect the valuable free text data collected via CDP disclosures. This data in its qualitative state is difficult to compare at scale but needs to be compared and assessed with companies’ and cities’ peers, suppliers, and neighbors to address water and climate risk challenges.

1. Utilize a subset of freeform response features to inspect from the water security and/or climate risk responses.

2. Apply your choice of NLP application (Sentiment Analysis, Classification, Question Answering, etc.) to the subset of data you chose. Explain the metrics you used to benchmark the success of your algorithm.

3. Visualize your data set by combining it with third party geospatial data and applying graph analytics to establish risk patterns.

[For inspiration review the CDP 2018 Global Water Report here: https://www.cdp.net/en/research/global-reports/global-water-report-2018]

[See the Hackathon Overview Deck linked above for more information on the challenges.]

Challenge 3: Renewable Energy Demand Projects

Currently only just over 10% of U.S. electricity comes from new renewable sources and this needs to ramp up quickly to meet growing renewable energy targets and demand. By focusing on matching renewable energy supply to meet demand signals in a given metropolitan area or electric utility’s service area, we can help green the electric grid more quickly.

1. Using CDP and other external datasets, identify where the current city and corporate demands/commitments are for renewable energy in the US and where they will be by 2023.

2. Select a city or region and build a predictive model that shows how that city will make progress towards meeting renewable energy supply and demand by 2023.

3. Explain the assumptions you’ve made to build your model and comment on the additional information needed to scale your model beyond 2023.

[See the Hackathon Overview Deck linked above for more information on the challenges.]