Smart city uses analytics and IoT to predict and manage flood events

The Town of Cary, NC, teams up with SAS and Microsoft Azure to protect citizens from flooding, safeguard watersheds and support environmentally sound development

Nice neighborhoods, top-rated schools and plentiful parks – all reasons why many people flock to the Town of Cary, NC. Over the last 25 years, Cary has tripled in size, with the current population exceeding 175,000. This influx of residents has led to a boom in new housing, shopping centers, businesses and entertainment venues.

As Cary’s population continues to grow, this smart city with small-town appeal has a team of people dedicated to using cutting-edge technology for the benefit of the community. This includes using the Internet of Things (IoT) to ensure that citizens are better protected from flooding events, new development projects are properly vetted so they won’t lead to flooding, and local watersheds are safeguarded.

“We want the Town of Cary to be the place where everyone wants to live, work and play,” says Nicole Raimundo Coughlin, Chief Information Officer for the Town of Cary. “Smart cities and IoT technology are going to drive the future of the way we operate within the municipality, and I’m excited to be part of a community that embraces data.”

From reactive to proactive to predictive

Stormwater, which runs off impervious surfaces like roads, parking lots, sidewalks and buildings, is a key initiative for the town. “One of the unique things about Cary is that we’re at the top of two river basins,” says Terry Yates, Smart Cities and IT Project Manager. “The water we get affects other communities in the state. This means our stormwater initiatives impact Cary citizens as well as citizens in nearby municipalities.”

Traditionally, Cary’s response to flooding was manual and reactive. “Before the Town of Cary had IoT sensors, the citizens were the sensors,” Yates says. “But a town should know what’s going on before a citizen calls in.”

“During storm events, we didn’t have visibility into river levels or how quickly water was rising – we didn’t have the technology to observe those dynamics,” Raimundo Coughlin adds. “Typically, we’d receive phone calls from citizens notifying us about a flooding incident. Then the town staff would reach out to Public Works or Public Safety personnel, dispatching them to the scene to put up barricades, close roads, redirect traffic and respond to emergencies.”

Raimundo Coughlin knew there had to be a better way to serve the community in flood-prone areas. “The Town of Cary is viewed as a leader in innovative city solutions. We’re committed to using smart technology and data to optimize city operations and improve quality of life, so it was time to apply that technology to predict and manage flood events.”

One of the town’s goals was to improve situational awareness using analytics and IoT. Prior to implementing its flood-prediction solution, the town’s data resided in multiple disparate systems. “We couldn’t get a clear picture,” Yates says. “We’d have to go to different places and manually coordinate a response. Nothing was automated.”

The Town of Cary wanted to eliminate data silos so departments and decision makers could benefit from a broader view of information they previously couldn’t see.

“We needed a scalable solution to move from reactive to proactive and ultimately predictive,” Raimundo Coughlin says. “In a flooding scenario, time is of the essence. Imagine there’s an occupied car that’s being swept away by water – every second counts. Ultimately, Cary’s flood-prediction system is centered around the safety of our community.”

SAS and Microsoft collaboration leads to optimal solution

Leaders from the Town of Cary met with representatives from SAS and Microsoft to plan out a new flood prediction solution. “We reached out to our technology partners and used our government municipality space as a lab to figure out what everything meant, whether it was security, processes, integration or data capture,” Raimundo Coughlin says. “This big collaboration was a win-win opportunity where we figured everything out together, and it was the best decision we ever made.”

Invigorated from these sessions, the Town of Cary gained a solid vision for the future and moved forward with the SAS flood incident prediction and preparedness solution powered by Microsoft Azure IoT.

“Having data from various systems unified in one place for a holistic view is huge,” Yates says. “There will be no more ‘Hey, we've got this information in an Excel spreadsheet, and we’ve got that information over in the public safety system, and we have this other information with the stormwater sensors.’ Having all the data pulled together in one place – and the ability to see and analyze everything together – is a major accomplishment for the town.”

This solution builds on the town’s cloud-first strategy. “The cloud enables folks who are in the field to have the information they need right when they need it,” Yates says. “And we can make architecture changes quickly and easily, which would be much more difficult with an on-prem installation.”

“We're very fortunate to have strong relationships with SAS and Microsoft because they’re not just technology vendors – they’re true partners who care about the work we’re doing, and they understand our vision and the impact it has in Cary and beyond,” Raimundo Coughlin says. “For us, this union is a dream come true.”

Analytics and IoT in action

The Stormwater Division is one of the main departments within the Town of Cary that relies on SAS Analytics for IoT. The stormwater staff receive data from solar-powered, cellular-enabled sensors that measure water height and depth, current flow and rainfall. Data from the sensors is uploaded to the cloud and combined with weather data. The Stormwater Division then can analyze the data and share information with other departments, such as Public Works, Public Safety and Information Technology.

“Our IoT pilot project is the Walnut Creek Watershed located in downtown Cary,” says Matt Flynn, Stormwater Development Manager. “We placed rain gauges and stream sensors in multiple areas throughout the watershed, and now we can see what is actually happening on the ground relative to what we are modeling and suggesting.”

Using the sensors, Azure Maps Weather forecast data and SAS Event Stream Processing, the town expects to increase situational awareness of rising stream levels, better predict where flooding might occur, deliver advanced warnings and improve emergency response through automation.

“The rain gauges give us a sense of how much variability there is in storm events,” Flynn says. “We can see whether it’s raining substantially in one location but not much in another. Then the stream sensors give us information about the waterflow that is in the stream. Having the ground truthing gives us the information to go back to our modeling and ask, ‘Is this what the model originally predicted, or is there new information that is giving us different outcomes?’”

The models are dynamic – they continually receive new information, which helps the town understand what’s actually happening on the ground and how to improve it. Staff members are able to pinpoint specific areas to focus on.

With predictive analytics, the Town of Cary can go beyond learning what happened before – and why – to discovering insights that will help it better prepare for future flooding events. Staff members can use data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on the data, and send out alerts in advance.

Additionally, SAS Visual Analytics provides the town an interactive dashboard, reports, business intelligence and analytics – combining traditional data with location data for analysis in a geographical context. Users who are out in the field have the ability to access the stormwater IoT dashboard from a mobile device. The information they need is at their fingertips, even when they’re not in the office.

“Having visual information is especially important because you’re able to easily absorb it,” Raimundo Coughlin says. “You don’t need to be a data scientist to understand the results. You’re not digging through a bunch of data, trying to find something. Instead, you're looking at images that actually show you the status of something.”

Using the Microsoft Azure IoT Hub, the town brings in sensor data and connects it to SAS solutions on SAS Viya. “With those pieces in place, we can see all the data flowing through the systems in real time, and we know where we need to make adjustments, like moving certain sensors to more stable locations,” Raimundo Coughlin says. “The scalability of Azure IoT platform became a vital component of our architecture, and SAS provided the industry-leading analytics that could deliver insight from both real-time and historical data.”

With reliable data streaming in, the Town of Cary can take steps to better predict the actual impacts of flooding. As various amounts of rain fall in different locations, the town can see the timing that houses, businesses, roads and other structures will be affected. Modeling can help staff better prepare and plan a course of action.

Water-flow monitoring via machine learning

While the Town of Cary is just beginning to explore the role artificial intelligence (AI) can play in flood prediction and management, staff members have already seen some promising developments.

“Right now we’re in the early stages of using AI, and we’re excited about growing in this area,” Yates says. “One example of machine learning is when the town experiences a rain event – the sensors recognize there’s a condition changing and subsequently start sending data faster, without any kind of human intervention.”

AI also plays an essential role when it comes to detecting issues with sensors. “We have a feedback loop that’s monitoring battery levels and power levels, and that system is constantly learning,” Yates says. “When an anomaly is detected, we can send it over to our work order management system to get someone to take a look. This demonstrates why it’s so important for us to have the ability to visualize water flow and actually see what’s happening.”

Better outcomes for citizens and the environment

Stormwater doesn’t just affect residents, buildings and roadways – it can have significant consequences on the environment. Untreated stormwater, which feeds into sewer systems and ditches, often carries debris, sediment, bacteria and chemicals that make their way into nearby streams, rivers and lakes. This subsequently degrades water quality and harms aquatic life.

“Stormwater management is a mix of engineering and non-engineering practices,” Flynn explains. “It’s choosing to develop certain areas that won’t exacerbate water quality issues, while not developing other areas that could create or worsen downstream flooding.”

The Stormwater Division plans to replicate what it did with Walnut Creek in all of Cary’s watersheds. The next phase of the project will be placing rain gauges and stream sensors in Swift Creek. “We know that Swift Creek going to tell a totally different story, and we’re eager to see what the data will say,” Flynn says.

The Town of Cary prides itself on taking care of its citizens and the environment, and it’s now in a better place to protect both from the dangers of floodwater.

“The beauty of SAS Analytics for IoT is that it helps the town council make decisions for the greater good of the community, while showing the logic behind those decisions,” Flynn says.

“The data helps us validate why we choose to develop new neighborhoods and businesses in certain places while avoiding others, or why we decide to acquire property in certain areas to create wildlife corridors. Ultimately, SAS gives us credibility, and in government that’s invaluable.”

Looking out for the well-being of others

Director Danna Widmar of the Town Manager’s Office is responsible for the Stormwater Division, along with several other teams dedicated to helping citizens. These teams use analytics to examine, reassess and determine what the town could do differently to make the biggest positive impact. It provides citizens with reliable information so they can make informed decisions regarding their homes, property and well-being.

“While the town can’t control Mother Nature, we will do everything within our power to assist citizens with the circumstances that occur on the ground,” Widmar says. “Flooding creates uncertainty, and that uncertainty is a burden that people carry. We’re here to combat that burden.”

Through a combination of data analysis, empathy and action, the Special Projects team works to understand the difficulties people experience with flooding issues and stormwater issues.

“Building dynamic models and using IoT devices in the field gives us greater certainty in predicting what’s actually going to happen, which is important to us and the citizens of Cary,” Widmar says. “The reality of the situation is there’s no way to eradicate flooding – our work is to mitigate its impacts. Sometimes that’s a very difficult thing, and it’s essential we understand the perspectives that our citizens bring to the table. When we work together, we have a much better chance of coming up with solutions that benefit the community and those who call it home.”

The impact of flooding extends beyond the physical cleanup and repair. Flooding can be an incredibly emotional experience. “If your home or property are damaged in a flood, there’s so much to consider,” Widmar says. “You’re dealing with the physical aftermath of the event, as well as the stress and sadness from losing things that were valuable and sentimental to you.

“Because the work we do is analytically driven, we’re able to make objective, unemotional decisions. While we know we’ll never be able to solve flooding, we can use data to improve the quality of life for members of the community. When a decision or action is data-driven – and you have the analytics to support it – then it makes sense to our town council and our citizens.”

A brighter future for Cary and beyond

The reach of the town’s analysis extends beyond Cary. The town shares its sensor data with regional partners, including the State of North Carolina Emergency Management, Wake County, the City of Raleigh, and the towns of Apex, Morrisville and Holly Springs.

“Stormwater flooding is not just a Cary issue or a North Carolina issue – it’s a worldwide issue,” Raimundo Coughlin says. “We’ve put a lot of thought and planning into the way we address flooding, and our town has an amazing opportunity to share our work and findings with other cities and states, demonstrating how communities can use analytics and IoT to better prepare for and respond to flooding.”

Regardless of their individual roles, staff at the Town of Cary share a common priority – the safety and well-being of the community. “Thanks to our integrated SAS and Microsoft Azure IoT solution, we have reliable data and the tools to easily interpret it,” Raimundo Coughlin says. “This empowers the town to make the best decisions possible regarding flooding, which, in turn, helps us better protect our citizens. We’re not just analyzing numbers – we’re looking out for our families, our friends and neighbors.”