Town of Cary innovates flood prediction with IoT

已于 六月 16, 2020 发布

Principal Program Manager, Azure IoT Business Acceleration

This post was co-authored by Daniel Sumner, Worldwide Industry Director, Government—Smart Infrastructure at Microsoft.

According to Flood Safety, flooding is the most common type of natural disaster worldwide. It affects tens of millions of people around the world each year and causes, on average, more than $200 billion in damages. Many communities face flood-related challenges, and the Town of Cary in North Carolina, United States, is no different. Its flood-prone areas are affected by heavy rains, which are often exacerbated by the yearly Atlantic hurricane season. When the town sees excessive rainfall, its personnel often find themselves scrambling to address overflowing stormwater systems, but even a burst water main can create a spontaneous flood event.

Town of Cary parking lot during a flood. Several cars parked in three-foot high flood water.

Town of Cary parking lot during a flood event.

As a leader in innovative city solutions, the Town of Cary was already committed to using smart technology, data, and analytics to optimize city functions, drive economic growth, and improve the quality of life. Chief Information Officer, Nicole Raimundo, Smart City Strategist, Terry Yates, and Stormwater Operations Manager, Billy Lee, saw another opportunity: use technology to predict and manage flood events.

Envisioning a flood prediction solution

In October 2019, Cary’s leaders met with partners Microsoft and SAS, IoT division, to envision a new solution. The team started by assessing the current situation.

During storm events, Cary had no visibility into the river levels or how quickly the water was rising. Traditionally, the town relied on citizens to alert them of floods through phone calls, text messages, and other means. The town staff processed these requests manually dispatching public work personnel to erect barriers and close roads and first responders to emergencies.

The team came away with a vision for building a flood prediction system leveraging Azure IoT and SAS Analytics for IoT. Raimundo explained the need for the change.

“We felt strongly that the existing system wasn’t serving citizens in flood-prone areas well. We knew we needed a scalable solution to get us from reactive to proactive and ultimately predictive. The scalability of Azure IoT platform became a critical component of our IoT architecture. In addition, we required a robust set of analytical tools that could deliver insight from both real-time and historical data and SAS Analytics for IoT offered that.” —Nicole Raimundo, Chief Information Officer, Town of Cary

“There are thousands of cities that are similar to the Town of Cary that are looking to deploy solutions to solve urban issues such as flooding. Leveraging the Azure IoT platform and SAS Analytics for IoT these cities can move from being reactive to proactive and, ultimately, predictive in a cost-effective, scalable manner.” —Daniel Sumner, Worldwide Industry Director, Government—Smart Infrastructure at Microsoft

Defining project goals

Cary, Microsoft, and SAS agreed to several project goals outlined below.

  • Improve the situational awareness of town staff.
  • Automate stormwater personnel notifications and work order generation.
  • Alert citizens of flooding events.
  • Provide data to downstream regional and state entities.
  • Analyze captured data and predict future flood events.

A key requirement for the Town of Cary was that their new flood prediction system needed to integrate with existing business systems. These included using the SAS Visual Analytics dashboard integrated with ArcGIS for real-time visualization, Salesforce for alerts, automated notifications and work orders, and data sharing for regional partner response systems.

“The Azure IoT platform has been a critical piece of our technology ecosystem and accelerates our ability to scale.” —Terry Yates, Smart City Strategist, Town of Cary

Through a series of work sessions with the partners in February 2020, the team created a project plan and system architecture. Then the implementation work began.

The Town of Cary team working in a conference room with Microsoft and SAS resources.

Town of Cary working session with Microsoft and SAS resources.

Implementing the solution

The Town of Cary installed water level sensors at various points along the Walnut Creek stream basin and rain gauges at several Town of Cary owned facilities.

Map of the Walnut River area showing where water sensors and rain gauge sensors were placed.

Water sensors were placed at strategic locations.

Below are highlights of how the solution was built.

  • Microsoft Azure IoT Hub enabled a highly secure and reliable communication to ingest stormwater levels over an Firstnet LTE wireless connection. The team used Azure IoT Hub to provision, authenticate, and manage the two-way communication to the sensors.
  • SAS Analytics for IoT combined streaming sensors or gauges and weather data for real-time scoring, dashboarding, and historical reporting.
  • SAS Visual Analytics provided interactive dashboard, reports, business intelligence, and analytics. The dashboard is integrated with ESRI ArcGIS for additional geographic analysis and data visualization.
  • Microsoft Azure Logic Apps seamlessly integrated with Salesforce and other third-party applications.
  • Microsoft Azure Synapse Analytics provides data warehousing for Big Data analytics.

Evaluating results

The solution’s initial phase has been running for several months with positive results.

  • Town staff can now visualize flooding events in real-time.
  • Stormwater personnel receive notifications and can generate work orders automatically.
  • A mechanism has been established to share data with regional partners.

“We’re still connecting some of the dots, but we’re already seeing real benefits in the automation of formerly manual processes. Previously, we might get a call from a citizen, which would cause us to dispatch public works or emergency services depending on the type of flooding. Now the data triggers alerts that automatically notify stormwater personnel, who can react and address the flooded areas. It’s much more efficient and could ultimately save lives.” —Nicole Raimundo, Chief Information Officer, Town of Cary

Lee explained how exciting it is to be able to visualize water flow and using the SAS Visual Analytics dashboard which is fully integrated with the ESRI ArcGIS.

“Now we can see a storm event in real time. We can pull up the dashboard and see how much rain we’re getting. We can see the stream levels rising and share this data with our regional partners. It’s amazing to see the data in real-time.” —Billy Lee, Stormwater Operations Manager, Town of Cary

Town of Cary storm water IoT dashboard.

Town of Cary storm water IoT dashboard.

Applying analytics

As the Atlantic region nears the peak of hurricane season, Cary’s leaders are looking forward to better predicting potential flood events. Leveraging SAS Analytics for IoT and SAS Event Stream Processing (ESP), the Town of Cary has enhanced their ability to acquire and manage new data from Azure IoT, generate and deploy predictive models, manage the lifecycle of those models over time, and achieve greater insight they can take action on.

Using Microsoft Azure IoT with the capabilities to integrate the water sensor data, Accuweather data from Azure Maps, and SAS analytics we are able to create a digital twin of the watershed. This allows the Town of Cary to be proactive in addressing floodwater issues so action can be taken ahead of the storm or flooding event.” —Brad Klenz, Distinguished IoT Analytics Architect, SAS

In the case of the flood detection and management solution, the Town of Cary can better identify anomalies, such as rising water, through the integration of weather forecasting data, real-time sensor data measuring water and rain levels to deliver advanced warnings and future predictions of flooding events both within the Town of Cary and downstream to surrounding municipalities.

“Cary sits on top of several rain basins. We will now be able predict flooding and share this information with our regional neighbors. This data and predictability will have a huge economic impact, not just in the Town of Cary, but for many municipalities, including local businesses and citizens, downstream.” —Nicole Raimundo, Chief Information Officer, Town of Cary

Advice to other cities

The Town of Cary has implemented a series of smart city initiatives, and its flood prediction solution shows amazing promise. What advice would Raimundo and Yates provide to other cities looking to implement similar projects?

“It’s really about selecting the right partners that understands your platform strategy vision for building solutions on a future-proof scalable architecture and that offer a flexible and open set of tools.” —Nicole Raimundo, Chief Information Officer, Town of Cary

Yates encouraged his peers to get the buy-in of all stakeholders.

“Include all departments, all subject matter experts in the digital transformation process and especially people working out in the field. You’ll need everyone’s buy-in and participation to be successful.” —Terry Yates, Smart City Program Strategist, Town of Cary

Next steps

Learn more about Azure IoT, SAS Analytics for IoT, and Microsoft for smart cities.