Using Crowdsensing for improved climate adaption

This year Codepan attended the Climathon as a team to see how we can apply our technical expertise and develop, in under 24hrs, a viable, future-focused product for a more sustainable future.

The Climathon, as the name suggests, is a climate-focused hackathon, run annually by EIT Climate-KIC, a European knowledge and innovation community, working towards a prosperous, inclusive, climate-resilient society founded on a circular, zero-carbon economy. It takes place simultaneously in over 100 cities worldwide. With over 5000 participants seeking to develop tangible solutions for local problems, it is the largest hackathon of its kind.

The problem?

As the climate changes, extreme weather events are becoming more frequent. How can we use open or crowd-sourced data, bypassing data silos, to better understand and build early-warning systems for these events?

Ideation workshop at the Climathon

Meanwhile, there are around 108 million smartphones in Germany and 77% of the population lives in urban areas. There are 38 million active social media users giving constant updates about the status quo.

Our solution: CrowdWeather.

CrowdWeather turns participating citizens smartphones, into a global network of smart weather stations. Modern smartphones contain thermometers, barometers, hygrometers and other sensors, all of which can be exploited for greater data integrity measurements within a city.

Design model of the CrowdWeather app [Panels: improved alert systems / MyStation metrics / localised weather risk maps]

Combining this with data scraped from social media networks and openly available data provided by government institutions, CrowdWeather improves the reliability of forecasts by obtaining more real-time, localised environmental data, without the need for any new sensors being built, or maintained.

Will (Data scientist) & Joel (Full-stack developer)

In less than one day, we developed a design model for the app and included some initial data science insights, using historical data crawled from GoogleTrends and correlating this with historical measurements of rainfall across different districts in Berlin.

Combining crowd-sourced data with social media information
A potential application of now-casting using google searches for improved weather updates

Our solution was developed with insights and feedback from researchers at the Potsdam Institute of Climate Impact Research as well as BerlinWasserBetrieb.

Urban heat island effect in Berlin

Cities of the future are not only susceptible to more flood events, but also the Urban Heat Island Effect, which contributes to global warming and also is a health risk to civilians. CrowdWeather takes advantage of mobile phone thermometers to map hot spots in the city to advise tourists/citizens to avoid heat spots during heat-waves and to provide planners with a dynamic reference to design systems to cool these areas.

Product Vision

Our vision is to create a city-wide network of citizen participation increasing Berlin’s resilience and improving climate adaptation. CrowdWeather is an open-source AI platform for predicting extreme weather events.

All participants of the extreme weather challenge celebrating (including some our team: Katherina, Philipp, Joel, Will, Ido and Sam)

The 2019 Climathon was a great success for the global community of climate action designers, showing what great minds can create given the right environment. We had great fun participating and it was also a great opportunity for our team to bond and work on a climate project together.

Interested in hearing more about CrowdWeather or about what Codepan can build for you? Drop us a message

We are digital architects and technology pioneers driving the sustainability revolution [Berlin, DE]