cholu6768 / air-quality-amsterdam-

Developed an analysis to improve the air quality in Stadhouderskade, Amsterdam

World Data League 2022

Executive Summary Template

This executive summary is one of the mandatory deliverables when you submit your solution. Its structure follows the WDL evaluation criteria and it has dedicated sections where you should add information. Make sure your executive summary covers all the sections since it will be an integral part of the Insights Report and your evaluation. Make sure your content is relevant and straight to the point. There is no need to reach the maximum number of words.

Instructions:

1. 🧱 Create a separate copy of this template and do not change the predefined structure
2. 👥 Fill in the Authors section with the name of each team member
3. ✏️ Write your executive summary - make sure you write to a non-technical crowd. You can refer to images that are in the Submission Notebook.
4. 📄 Fill in all the text sections
5. 🗑️ Remove this section (‘Executive Summary Template’) and any instructions inside other sections
6. ⬆️ Upload the .md file to the submission platform.

🎯 Challenge

Air Quality Prediction in Busy Streets

👥 Authors

Include all the authors that have worked on this submission. It is not obligatory to include all the team members if not everyone has worked on it. This will not impact the evaluation in any way.

• TinWan Ng
• Fernando Chavez
• Kirti T.

✨ Introduction (250 words max)

Provide a contextualization of the problem, together with an estimation of its size using real numbers and references. Write here

🔢 Data (250 words max)

Explain what data you used (both provided by WDL and external) and improvements you suggest to those datasets. Explain how those improvements would lead to a better solution.

We used the air quality data provided by World Data League and weather data from xxxx website.

🧮 Methods and Techniques (250 words max)

Tell us what methods and algorithms you used and the results you obtained. Write here

💡 Main Insights (300 words max)

1. Between 2014 and 2022, the benzene level declined, moving from a moderate level (0.972 - 9.75 ug/m3) to an acceptable limit of below 0.975 ug/m3. If the concentration rises to a moderate level, poisoning may occur, with symptoms including speech problems, headaches, dizziness, sleeplessness, nausea, paraesthesia in the hands and feet, and fatigue; however, if the concentration rises to a high level, the risk of death rises as well.

2. All other pollutants are still under the permissible limits, and their values have decreased slightly from 2014 to 2022. Decrease percentages are mentioned below for all pollutants:

• Benzene - 34.5%
• Toluene - 71%
• Xylene - 81%
• NO2 - 37.8%
• PM10 - 11.5%
• PM25 - 24%

3. During the summer months of May to August (Summer), the pollutant value decreases, and the air quality index increases. However, during the month of December, the pollution levels skyrocket.

4. Pollution levels for NO2 are lower in the early mornings when compared to the entire day where they are consistent. For other pollutants, there is no such pattern observed.

🛠️ Product

Definition

Define in one sentence what product(s) could be built out of the code you produced.

A web application that shows visualizations of the pollutant values over time and informs about the prediction (1 week ahead) of the air quality including a warning system.

Users

Describe who would be the users of your product and for what purpose would they use it.

People in charge of the project could use it to see how the pollutant values have changed over the years, months and time of the day. They could also see the prediction of the values up to 1 week ahead of time.

Activities

Describe what features your product has. Example:

• Predicts the air quality up to one week ahead of time
• Visualizes the values for every pollutant over a timeframe of years, months and changes over one day.

Output

Describe what the product outputs to the users and how it does that. You can add mockups and/or visualisations.

Example: Location of the accident on a map and suggest the fastest route from the dispatch centre.

🌍 Social Impact Measurement

Outcome

If the outputs are your immediate results, describe your long-term results. What do you want your product to achieve? What ''good'' are you creating?

Example: To decrease response time from dispatchers so that people in urgent need receive help faster.

The product should be publicly available so that everyone can see and interact with the measured pollutant values over the years, what might be causing them, prediction of the pollutant values over time.

By making this product publicly available, the community would get a bigger notion and could also take part in the change.

Impact Metrics

From the outcome, define 2 to 4 metrics that measure if you are achieving that outcome or not. Example:

• Average Dispatch Time
• Average Distance from Accident Location and Dispatch Center

Impact Measurement

Since you cannot wait to see the impact of your product, estimate it. You can do that by either using the estimations/predictions of your model, market research, products from proxy industries and/or similar locations, etc.

Example:

• Based on model predictions: Our model estimates a decrease of 6 minutes of the average dispatch time and a decrease of the average distance of 200 meters
• Based on proxy products: Similar studies in other cities show that the dispatch time can be decreased by as much as 13 minutes, depending on the traffic intensity of that city.