Why is Raspberry Pi perfect for building flood resilience?

Creator meets user 

Raspberry Pi was founded by Rob Mullins and five other friends in 2009 at Cambridge University. Rob and Eben Upton (now CEO of Raspberry Pi trading) met to discuss “how applicants for computer science had fallen sharply and how those applying had less experience than in the past. The solution, we thought, was to build a low-cost computer. The idea was that this would be something that children could own, experiment and create with and build into exciting projects."

Since then, more than 15million Raspberry Pi computers have been manufactured and it has become the go-to technology for creating low-cost, yet powerful, solutions to local problems.

Miguel Arestegui and his team have used the technology to adapt and improve flood early warning systems in Peru

“As part of the Zurich Flood Resilience Alliance we’re working in communities [in Chosica, Peru] where we have the problem not only of floods but also of rapid debris flow”. 

Without enough warning, communities cannot escape the danger. Although national warning systems do exist, Miguel explains, "There is a strong distrust of flood forecasts, because we have a serious lack of historical data in this area". An effective local solution was clearly needed.

Miguel’s team discovered that the “Raspberry Pi Foundation has this weather station kit  for high schools, and that gave us an idea: what if we tackle this need…[by] adapting or developing this sort of early warning system?” 

Miguel and Practical Action then worked with the community and local government to implement a warning system controlled by monitoring stations based on Raspberry Pi. The micro-computers receive and process information on rainfall, soil moisture and river water levels, and take pictures. This information then feeds into platforms that issue alerts.

An ‘end-to-end’ early warning system composed of four components, including monitoring which in Chosica is controlled by Raspberry Pi. Source: Practical Action

From toy box to tool box

So what is it about the Raspberry Pi that makes it the perfect flood resilience tool?

It’s open-source 

“The fact that the board is used by so many people means it’s become a standard component” says Rob. Plus, because all those users are creating projects with open-source code “other people can build on them and improve them”. This means there’s a community of people and experience to kick-start new projects.    

Miguel’s experience demonstrates this advantage: “We completely constructed this [early warning system] with the community. The open-source code was the building block that helped us complete this in […] just over one year. It would have been impossible if we were working on our own.”     

It’s adaptable 

Because users have full access to Raspberry Pi codes without commercial constraints, the technology can be tailored perfectly to fit need. “People are able to use computers as tools” says Rob, “they’re able to produce the solutions themselves rather than having to go to someone else to provide the implementation. This stimulates local solutions to local problems.”  

For example, in Chosica the previous early warning system was controlled by nation-wide commercially-owned software, which made local-scale changes impossible. But Miguel explains that because the new system was based on Raspberry Pi adaptations could be made based on local knowledge, for example “to take data more often than what technical studies would suggest. This was later found to be necessary based on the short lead time for these rapid events, stressing the importance of local memory in data scarce regions. The fact that these technologies can be locally adapted makes them good for building resilience, which goes way beyond isolated preparedness measures.”

It’s low-cost but not low-tech

“Previously, low-cost implied low-tech” says Miguel. To have both high-tech and low-cost “is providing a new platform that could help link the gap between local needs in developing countries and the usual high cost of equipment that hinders National Scientific Institutions to address those needs”. 

Rob agrees, he has seen that “there is enormous scope to…replicate an expensive and very specialised system using something like Raspberry Pi to produce something that is technically almost as good, but using a very low-cost solution.”  

The future of risk, resilience and Raspberry Pi

How do our experts think the technology will change in the future, and how could this make an even better tool in the fight against climate change? 

Better hardware

Rob, as the hardware expert, thinks the next few years will bring “ultra-low power computers that can be used in these monitoring applications. Also computers that eventually just biodegrade and don’t have the impact on the environment that they do today.”

Better connectivity

Miguel sees a future where more and more of us are connected to the internet. “Right now there are some constraints with connectivity that I think are going to start changing quite rapidly. [This] is going to provide so many tools for people in vulnerable situations”. 

Changing the way people can share their own knowledge will help them cope with climate change. According to Miguel, “in climate vulnerable areas there is a critical problem in the lack of connection between the impacts of climate change and the amount of data you have in those areas. I think that these technologies can help the role of local information that communities themselves can provide to address this gap.” 

Better networks 

Rob says “even though the community is very strong I still think there are opportunities to build better networks, for example between universities in different countries”.

But Miguel predicts better networks may also need to exist at higher levels. “I am curious to see if we arrive at some sort of standardisation of these open-source and decentralised initiatives. It is useful to adapt development to local contexts but I think to scale-up these initiatives and make them reliable and robust enough to take to high-level discussions, a level of standardisation is needed.”