RISE Innovator Spotlight
Juliette Murphy: FloodMapp co-founder and CEO
Navigating flooded streets is not just an inconvenience. It can be life threatening.
Last year alone, nearly a third of the 139 flood fatalities reported by the National Weather Service as of October occurred in vehicles. FloodMapp, a Brisbane, Australia-based startup and RISE’s 2020 Urban Mobility Resilience Challenge winner, aims to make avoiding watery roadway hazards easier.
“We want to see a world where no lives are lost in flood events, where people get home safe to their families and loved ones, always,” says Juliette Murphy, the company’s 34-year-old co-founder and CEO.
After witnessing the impacts of devastating floods in both Australia and Canada, the Brisbane native teamed up with Ryan Prosser to develop integrative technology that enables real-time, street level flood warnings, improving upon the more broad-based and generalized area alerts available today.
“With my flood modeling background, I saw the experience play out not just as a resident, but from an engineering perspective,” says Murphy of those 2011 and 2013 flood events. She could see the technical gaps, and the dots that needed to be connected, she adds.
“I felt a real purpose and responsibility that it was me who needed to step up and solve this problem.”
In January, drivers in Norfolk, Virginia, became the first in the world to participate in the next evolution of that technology as part of FloodMapp’s new pilot program: the integration of their NowCast flood forecast technology with the Waze navigation app. The project launched days before a significant flooding event hit the region, and for the first time, Waze users received audio alerts and pop-up icons warning them about flooded streets along their route to avoid car- and life-threatening situations.
News of the project attracted attention from more than 200 news outlets including the Associated Press, Weather Channel and Government Technology.
Since the launch, more than 3,500 drivers have used an in-app “thumbs up” feature to confirm the flood hazard alerts to help FloodMapp validate and refine their modeling capabilities. The ultimate goal is to reroute drivers around flooded roads — a feature currently in development.
“There is an old saying in physics: ‘All models are wrong, but some are useful,’” says Murphy. “With the community’s help, we hope to provide something that everyone will find really useful.”
Murphy says she was “overjoyed” to have FloodMapp named the winner for the RISE Urban Mobility Resilience Challenge and grateful for an opportunity to test and advance her team’s technology in Hampton Roads. “New technologies face a number of barriers to commercialization and widespread adoption, so I think the work that RISE is doing is critical to bring innovative resilience solutions to communities in need.”
The startup’s Waze pilot program is being conducted in partnership with the City of Norfolk. We caught up with Murphy to learn more about her and the emerging technology she has developed with her team at FloodMapp.
You’ve been a professional surface water engineer for more than a dozen years. Who or what inspired you to pursue a career in a STEM-related field?
Yes, I studied environmental engineering and went on to specialize in flooding and hydrology. My dad was a surveyor, so he had a strong background in math, science and mapping. I always found that really interesting. Then in high school, my career advisor suggested I explore engineering.
Tell me about how the Queensland, Australia, floods of 2011 and the Calgary, Canada, floods of 2013 inspired you to develop FloodMapp? What indelible images linger from those experiences?
Whilst in Brisbane, I personally experienced the devastation of the 2011 floods. I was working for Golder Associates, a global environmental consulting firm, as a water resource engineer on flood projects across Australia and Indonesia. In January that year, Brisbane was struck by a catastrophic, 1% probability flood that resulted in over 20,000 homes being inundated and causing more than $10 billion in damages across the state. Sadly, 35 lives were lost due to inadequate flood warnings. My friend’s house was engulfed by flood water, over the peak of her roof. She lost everything she owned. The image that sticks in my mind is seeing the house and all its contents upended, filled with mud and debris in the days following as we went to clean up. It was absolutely devastating. My friend had no idea the flood would cause such an impact. The warnings she received simply referred to flood height, which is not useful for most people.
In 2012, I transferred to the Golder Calgary (Canada) office for four years. In 2013, I experienced catastrophic flooding in Calgary with over $6 billion in damage and more than 75,000 evacuations. It was the opposite side of the world and yet the experience was eerily familiar. There were broad general warnings saying the river would peak at 13.9 meters. All my friends were texting me asking if they should evacuate, if they should move their car, as there was a lack of personalized, location specific warnings. Five of my friends came to stay in my apartment.
Seeing these events and the devastation firsthand is a big reason for founding FloodMapp. I couldn’t help feeling that the engineering industry needed to do more to provide better warnings and more location specific information.
How did you hear about RISE and what motivated you to apply for its Urban Mobility Resilience Challenge?
We heard about the Challenge from Sean Bohannon at Trade and Investment Queensland, who sent us the link. The Urban Mobility Resilience Challenge described the exact problem that FloodMapp set out to solve and was calling for technology that was essentially what we had developed.
So, we decided to apply as we thought it was highly relevant. But furthermore, because it was an opportunity to apply our technology in a way that could improve resilience for residents in Hampton Roads by helping them adapt to flooding and streamline urban mobility.
In essence, how does FloodMapp’s technology work?
FloodMapp’s groundbreaking, new flood model technology DASH is a real-time flood modeling system, purpose built for emergency management. DASH ingests incoming live sensor data from tidal, riverine and rainfall sensors as well as climate forecast models to rapidly predict flood inundation mapping at scale over a large area. DASH forecasts the inundation area and depth of an impending flood event down to an individual street address, with instantaneous updates to reflect changes in the real-time data inputs.
FloodMapp DASH covers three key types of flood events: coastal flooding, pluvial or flash flooding caused by heavy rainfall, and riverine flooding. DASH then feeds the live flood model results into mapping platforms commonly used by emergency managers and geographic information system (GIS) professionals, such as Esri’s ArcGIS online, and mapping apps including Waze.
What is the traditional method of flood forecasting and mapping? How does it compare to what your company is doing?
To produce a property specific flood inundation model at scale, three stages of modeling are required, all interdependent.
- Firstly, meteorology. Whilst no meteorology forecast will ever be perfect, there are many good atmospheric modeling solutions available.
- Secondly, hydrology. Traditionally, hydrology models are physics-based, calibrated to volumetric discharge, or a formula based on the amount of water released per second. But discharge cannot truly be measured. Flood sensors measure river height and discharge is later assumed based on limited hydrographer measurements and rating curves. Therefore, all models are built on multiple layers of uncertainty.
- Thirdly, tidal and storm surge modeling, which uses physics to predict coastal flood height.
Hydraulic modeling would be required to predict the location-specific impact of flooding to people, property and critical infrastructure, such as the exact flooded roads. These physics-based models are used to map the 1in 100-year floodplain, for example. They are manual and time consuming to set up, and can be computationally intensive — and costly. Therefore, traditionally, this step cannot be done in real time and is the biggest technical barrier to real-time flood intelligence.
At FloodMapp, we use data driven models, informed by physical hydrology and hydraulics. Our models introduce artificial intelligence and machine learning to break the barriers of speed, accuracy, computational cost and scale faced by traditional and manual models. In this way, we’re able to deliver street level flood intelligence in a time frame and with a level of accuracy that was never possible before.
What was the biggest challenge you faced as you were developing the technology?
Speed. We needed to build something that could read in current tide and rain conditions and return an accurate result in less than 15 minutes, and that had never been possible before.
Was there a particular moment when you realized you not only had a great idea, but you had something that worked?
Yes. In 2018, we were testing our technology in a simulation of a historical flood event. When our model matched up exactly with what was observed, we realized that what we had developed worked really well as a real-time flood modeling solution for disaster management.
Who are the end users of this technology?
The end users of our technology are emergency management and GIS professionals, across the public and private sectors. These users ingest our flooding data and application programming interface into their mapping platforms to inform their emergency response before, during and after the flood event.
What does this technology look like for residents of flood-prone areas? In other words, how will the average user benefit and use it?
Think of FloodMapp as the rainfall radar for flooding. Anyone can look at the National Weather Service rainfall maps and see whether it’s raining at the baseball field near their house; now you can use FloodMapp to see if there is water over the roads you would use to get there.
Floods can be devastating for local economies. How does your technology help local businesses?
Flooding can cause significant business interruption resulting from lost stock, flooded roads and lost access, and disruption to supply chain and logistics. These losses can result in businesses having to shut down, and subsequent unemployment and financial hardship. FloodMapp technology such as our ForeCast and NowCast products can help deliver businesses with property specific flood intelligence before and during a flood, which can help them move stock or relocate assets to prevent loss. It can also help identify and prevent impacts to the supply chain by rerouting delivery vehicles to prevent delays and shortages.
As winners of RISE’s Urban Mobility Resilience Challenge, you not only received funding but business mentorship. How was that helpful to your startup and what was your biggest takeaway from the experience?
Yes, as a winner of the RISE Urban Mobility Resilience Challenge, we had access to their Resilience Innovation Accelerator that they provide through George Washington University, including mentoring, which helped us test our assumptions and business hypothesis and get valuable feedback on our business plan. We have learned that there is a large market opportunity for FloodMapp’s solution with not only governments, but also utilities, transport and logistics companies, and insurers who are all looking to improve safety, prevent damage and reduce business interruption when it comes to flooding.
What changes did you make to your technology to align with the unique challenges the Hampton Roads region faces as a result of climate change?
Each city has its own unique challenges with flooding. Most major cities are built near some sort of waterbody, so it might be a storm surge from hurricanes, it might be a river breaking its banks, or just intense rainfall and an aging stormwater network that was never expected to cope with the volume of rain we are seeing these days. Sadly, the cities in Hampton Roads suffer from all three. We typically focus on river and coastal flooding as they cause the most damage, but for this project we had to build a custom flash flooding model to predict flooded roads, calibrated specifically to this region in an aim to get the most accurate results possible.
RISE provided Norfolk as a living laboratory for you to pilot your solutions. How does testing here help you scale your business?
We’ve learnt over the course of this project just how familiar the people of Hampton Roads are with flooding. Depending on where you live and where you work, you could experience flooding more than 12 times per year — which is staggering. That frequency of flooding provides a real opportunity to not only help this community, but to collect more on-the-ground, observable data to improve our models. We fully expect this to be an iterative process that will require repetition and adaptation, and we want to work closely with the community to provide them the best support that we can. Then our hope is to take what we learn here and use that to support other cities that are impacted by less frequent but more severe flooding.
When did testing begin and what kind of feedback do you hope to provide and/or what do you hope to learn?
Our early alpha testing began in 2021. We looked at historic events to compare our results with reported road closures and have been evaluating and improving the live maps as part of our beta testing since August. When we began the soft launch of the integration using Waze for Cities in late fall, we were been blown away by how accurate our models were.
Now that we’ve opened it up to the community, we want to encourage as many people as possible to download and use Waze to give them a tool to avoid flooded roads, especially as hurricane season approaches. We are also eager to get more feedback through the app, which helps us improve future predictions.
What aspects of FloodMapp’s technology are available now and where is it being used already?
We are a small but international company headquartered on the other side of the world in Brisbane, Australia. This Waze integration will be our first public-facing product, but we’ve been working with a range of organizations in a range of geographies. For example, we’ve been working closely with emergency services here in Australia. We worked around the clock during the recent catastrophic flooding to provide first responders not only extra lead time, but also more accurate warning areas to evacuate communities prior to dangerous flooding. We’ve also used the technology in North Carolina, Texas and Louisiana.
What elements of the technology are still in development and when do you foresee its future availability?
The core technology is and will always be in development. Flood modeling is a wicked problem. I believe people will want and deserve more and more accurate results, with better lead time. We all want a perfect weather forecast. The reason we don’t have it isn’t because people aren’t working on it, it’s only because it’s really hard. Therefore, at FloodMapp our multidisciplinary team of engineers, flood hydrologists, data scientists and software experts are committed to continual improvement and pushing the boundaries of what is possible. We’ll be launching in more locations as early as next year. We would love to see this traffic navigation capability offered to every city that needs it, as well as provide support to any business that suffers from flood related losses.
Last year, you were recognized for your work in artificial intelligence at the Australia and New Zealand Women in AI Awards, specifically in the category of mining. What does that award represent to you and how does it reflect on your efforts at FloodMapp?
I was incredibly humbled and honored to receive this award. The ANZ Women in AI awards aim to highlight the successful work of women in artificial intelligence and increase female representation and participation in AI. So, it represents a celebration of the incredible work being achieved by women in AI, which I hope can inspire other women in STEM to explore careers in AI. It also represents an incredible recognition of FloodMapp’s technology and how critical it is across multiple industries, including mining. Prior to founding FloodMapp, I worked with engineering consultants on some mine water management projects. I saw firsthand the catastrophic damage that flooding caused to mine sites. In 2011, a single flood event caused over $5 billion in damages to the mining sector. Some mine sites were forced to close, which had a devastating impact on communities, employment and the economy. FloodMapp is working with companies in the mining sector to better prepare and prevent this kind of economic damage from happening in the future.
How do you define resilience, and what is your vision of the future for our flood-prone, coastal communities?
I define resilience as the ability to rapidly adapt to a changing climate. To be properly prepared for what is to come, to endure and withstand catastrophic disasters, and emerge on the other side stronger and more prepared than ever before. Our core mission is to build a safer future. Our vision is to become the most trusted provider of flood forecast and real-time mapping data across the world, enabling us to make a global impact.
When our technology has been used to successfully evacuate communities ahead of a flood event, there is no other win that compares to that. Life is precious and our whole team is so passionate about improving safely in the moments that matter.