Announcing the winners of the 2023 Air Quality Global Hackathon

On December 6, 2023, Tech To The Rescue in collaboration with Amazon Web Services (AWS), organized an Air Quality Hackathon aimed at tackling one of the world's most pressing health and environmental challenges – air pollution. Participants were asked to use their creativity and technical skills to build cloud solutions that support local organizations and communities with access to clean air.

Air pollution is one of the most pressing health and environmental issues linked to 6.7 million premature deaths worldwide yearly (WHO). It most acutely affects populations in South, South-East Asia, and Sub-Saharan Africa.

More than 170 tech teams from 27 countries and 5 continents, with the help of over 50 technical mentors, used the latest cloud technology and Artificial Intelligence (AI) tools to build 33 solutions. The 3 winning solutions bridge the data, forecasting, monitoring, and information gap with the ultimate goal to clean up the air.

The Air Quality Hackathon, a three-day, global and virtual event, was aimed at finding innovative technical solutions to solve the most pressing air pollution challenges. Tech To The Rescue worked together with the world's leading non-profits and experts tackling air pollution to surface the key blockers they face. Hackathon participants then built technical prototypes to address those blockers and clean up the air for the betterment of the affected communities.

The hackathon was accompanied by the launch of a technical needs report and research “Fighting Air Pollution with Technological Innovations (Report 2023)” conducted by Tech To The Rescue. The report's findings include insights from interviews with leading non-profit organizations in the mentioned regions. These leaders are facing a multitude of technological challenges preventing them from scaling their proven interventions. Challenges include: access to air quality data collection, designing new sensors measuring the levels of pollution, creating open-source air quality databases, and crowdsourcing data. Furthermore, building awareness among the citizens by developing user-friendly interfaces to display air quality data in an attractive and actionable manner is critical. Another crucial component of the puzzle is to support advocacy work by highlighting the impact of new policies on the environment, people, and the economy.

“At AWS, we recognize the disproportionate impacts of air pollution on vulnerable communities around the world, the people and places least responsible for the problem; and are committed to using cloud technology to help mitigate this. We believe working together is the key to unlocking technology’s full potential in this space. Improving air quality is a complex issue that requires collaboration to both build solutions and make impactful air quality data available. Seeing the innovative prototypes built during the hackathon gives me great hope for our ability to combat this issue.”  Maggie Carter Director, Social Responsibility and Impact Amazon Web Services (AWS)

The 2023 Air Quality Global Hackathon was a call to action to the technical community to explore how cloud technology can point us towards solutions that support the goal of cleaning up the air. The hackathon judges and final jury, identified 7 finalists (1 per each challenge area) and among them 3 winners that maximize value and will scale, are simple to use for the local actors, broadly accessible and demonstrate creativity and strong alignment with the non-profit organization’s mission. Here are the winners:

Polish Smog Alert is an organization that conducts a campaign "See what you breathe. Change it". A seven two-meter high models of human lungs travels to 63 locations. The lungs "breathe," absorbing air pollutants that settle on the white matter covering the installations. The challenge for the tech team was to transfer the mobile lung experience to the digital space for the worldwide audience.

The Brainhub team has prepared and implemented a website and a mobile application written in React.js with a very suggestive image of breathing lungs. Depending on the location we choose, and actual real-time data from OpenWeather API, the image will show us the damage to the lungs appropriate to the air quality. In future development, the website will be able to display historical data. On top of this, the team invented a mobile game using the Kaboom game engine in which the user can become a brave lung that fights air pollution. The level of difficulty depends on the level of pollution. The application is easy to deploy on AWS or any other container-enabled environment.

The Thailand Clean Air Network is actively working to bring about meaningful change in air quality legislation. One significant hurdle the organization faces is the lack of appeal in legal language, which often alienates a broader audience, especially the youth.

The BlueRider.The software team prepared the AirVocacy, a digital platform that integrates live sensor data from over 1800 stations in more than 370 cities, providing real-time updates on air quality. Complementing this is a user-friendly chatbot powered by GPT 3.5 with an option to transition to GPT 4.0, which answers queries about air pollution and legislative measures. An AI meme generation tool utilizing DALL-E 3 to engage a younger audience creates shareable memes, simplifying complex legal and environmental concepts. Additionally, advanced legal language processing, employing Langchain & Qdrant, enhances the accessibility of legislative documents, ensuring precise answers in any language.

Afri-SET is working on a big project in West Africa to ensure affordable air quality sensors are used in places that usually get little attention. They want to create a smart database that can handle data from all kinds of sensors, no matter who makes them. Their goal is to make it easy for anyone with a sensor to join in and use their platform.

The solution crafted by The Camaraderie, named "Bedrock Mining the Air for Data Nuggets," introduces a novel approach to data processing. It utilizes a Language Model (LLM) prompt to generate Python codes automatically capable of handling raw data in various formats. This innovation significantly streamlines data engineering efforts, reducing the timeline from months to days. The system automatically generates code for each data format and seamlessly integrates with Amazon Bedrock, accessing cutting-edge LLMs via API. It incorporates human involvement to oversee data ingestion and LLM outputs for additional security measures. The solution also leverages AWS safety, security, and responsible AI features. Notably, it's easy to scale by using AWS Glue, Amazon Athena, and Amazon S3 as a data lake platform, offering support for various analytics services and facilitating easy extension to a production-ready system.

Congratulations to the winners and all the builders of the prototypes put forth during the Air Quality Hackathon. he winning solutions will receive financial support to be implemented in production.

If you would like to get involved and put your technical skills to work, we invite you to read Fighting Air Pollution with Technological Innovations (Report 2023), where non-profits fighting air pollution share their additional challenges to scale and start building for and with them. Reach out to Tech To The Rescue directly if you have inquiries, feedback, or ideas to utilize the Air Quality Hackathon's outcomes in supporting Access to Clean Air.

About Tech To The Rescue

Tech To The Rescue is a global program connecting technology companies with nonprofits around the world to solve the world's most pressing problems through technology. Since 2020, more than 1500 technology companies and over 1400nonprotis have joined the #techforgood movement, and our team has enabled more than 550 pro-bono technology projects, which help to boost the nonprofits' impact.

----

---