Floods are devastating. They are tearing apart communities, devastating neighborhoods, forcing the evacuation of thousands of people each year, and it can take years to recover – provided recovery is possible at all. The US government estimates that floods (excluding hurricanes and tropical storms) have caused an estimated $ 160 billion in damage and killed hundreds of people in the past few decades.
You’d think we should have a real-time model of where water is and where the world is going, with all these sensors on the ground and satellites in orbit. But we usually don’t do that, instead relying on outdated models that don’t take into account the possibilities of big data and big computing.
FloodMapp, a startup based in Brisbane, Australia, aims to wash out the old approaches to hydrology and predictive analytics and implement a much more modern approach to helping emergency managers and citizens know when the floods are coming – and what to do.
CEO and Co-Founder Juliette Murphy has spent a lifetime in the field of water resource engineering and has seen firsthand the severe destruction water can cause. In 2011, she saw her friend’s house flood amid a terrible flood. The “water went right over the top of her house,” she said. Two years later, she saw the same situation again in Calgary: flooding and fear as friends tried to decide if and how to evacuate.
These memories, and her own professional career, led her to think more about how she could develop better tools for disaster managers. She eventually teamed up with CTO and Co-Founder Ryan Prosser to build FloodMapp in 2018 and raised A $ 1.3 million along with a related grant.
The company’s premise is simple: we have the tools today to create real-time flooding models, but we chose not to use them. Water follows gravity. So if you know the topology of a place, you can predict where the water will flow. The challenge was that computing second order differential equations with high resolution remains computationally intensive.
Murphy and Prosser decided to avoid the traditional physics-based approach that has been popular in hydrology for decades with a completely data-based approach that uses widely used machine learning techniques to make these calculations much more palatable. “We’re doing top-down what used to be bottom-up,” said Murphy. “We really broke the speed limit.” This work led to the development of DASH, the startup’s real-time flood model.
FloodMapps modeling Brisbane river flooding. Image credit: FloodMapp
Unlike typical tech startups, FloodMapp doesn’t want to be its own independent platform. Instead, it interoperates with existing geographic information systems (GIS) such as ESRI’s ArcGIS by providing a data layer that can be combined with other data streams to provide situational awareness to emergency and rescue personnel. Customers pay a subscription fee to access FloodMapp’s data layer, and to date the company has partnered with Queensland Fire and Emergency Services in Australia and the cities of Norfolk and Virginia Beach, Virginia.
But the startup doesn’t just want to win the rescue services in the end. Any company with physical assets, from telecommunications and energy utilities to banks and retail chains with physical stores, could potentially be a customer of the product. In fact, FloodMapp is betting that the SEC will mandate further financial disclosures on climate change, which could result in … a flood of new business (I’m getting a flood pun, okay, I get one).
The FloodMapp team has grown from the original two founders to a number of engineers and salespeople. Image credit: FloodMapp
Murphy notes that “we are still in the early stages” and that the company will likely raise additional funding early next year as it survives this year’s flood season and brings on several new customers. She hopes that FloodMapp will ultimately “not only help people, but also help our country adapt and adapt to a changing climate”.