Hurricanes and AI: Savannah’s Unique Strategy for Severe Weather Management

Introduction

When Hurricane Dorian struck the southeastern portion of the United States, the City of Savannah used AI to track community feedback about their hurricane updates and emergency instructions. With this feedback data, the City then adjusted its messaging to ensure that vital safety information was presented effectively and correctly.

The Challenge

The City of Savannah, Georgia, needed to prepare residents for the severe weather conditions that were predicted to impact the City from Hurricane Dorian. To start, the Communications Department released information such as press releases and evacuation maps to inform the public on how they should prepare. The content was published through the City’s official website and social media channels, and then subsequently distributed by the local media. A major challenge for the  Communications Department was to understand what happened once the information was dispersed to the masses. Was the public properly informed or were there any miscommunication issues occurring? Time was of the essence and the City needed answers in order to mitigate damage and protect Savannah residents. 

The Zencity Solution

The City of Savannah’s Communications Department turned to Zencity’s AI technology to save time and better monitor their community’s reactions to the weather. In particular, the Department used the platform to track discourse varying on a variety of media sites and citizen-run Facebook pages and groups. This data helped them quickly understand which emergency weather resources worked better for their community. For example, the City produced two versions of evacuation maps for local media distribution to see which was more effective. The Zencity discourse analysis tool helped them identify which evacuation map was seen as more informative and if any element of the maps needed to be edited based on real-time community feedback. This qualitative, inclusive data helped the City to make accurate adjustments and promptly release critical safety information. Savannah’s city officials could also identify and rectify cases of life-threatening misinformation – specifically instances where residents stated on social media that they aren’t going to evacuate from flood-prone locations because they misunderstood the map. In the past, the city administrators often had to guess whether instructions and safety information were effectively reaching the masses and meeting the needs of the community. Zencity’s AI eliminated that guesswork and helped City Management navigate the severe weather with even better community-facing communication and resources.

Actionable, Data-based Insights

Zencity’s sentiment analysis tool was used to monitor community sentiment in the period before, during, and after the hurricane hit Savannah. The tool highlighted citizen satisfaction with how the public’s safety was managed by the City. The data also showed that the City’s proactive approach and communication towards the hurricane was directly aligned with the community’s needs.

The insights feed provided the Communications Department with a report that highlighted key takeaways on the online discourse surrounding the hurricane. The report itself combined real-time data collected by Zencity’s AI with qualitative analysis which focused on what the greater Savannah community was discussing online. For example, one interesting takeaway was that the majority of the community discourse took place either on local media sites or social media. This demonstrated the significance of these channels as an important source of information for the public. Furthermore,  the insights feed helped the City identify another potential resource, an Alderman’s social media page, that could be leveraged to communicate vital safety information to the community.

Access to “unofficial” citizen service requests, such as a resident’s Facebook post about a fallen tree, allowed the Communications Department to easily assemble resident feedback and turn them into official 311 requests in the City’s CRM system. This helped Savannah’s City Management optimize the documentation, delegation, and prioritization of service requests during Hurricane Dorian, while in conjunction, improve the administration’s overall awareness of the greater community’s needs.

Results

 

	The Communications Department adopted a data-informed, community-driven approach to severe weather management messaging.
	The City was able to quickly identify and respond to misinformation on public forums and pages.
	City Management used the platform as an unofficial service request system to identify in real-time where damage had occurred and allocate resources effectively.