Supporting 911 centers with non-emergency response solutions: An architecture guidance
Since February 1968, 911 centers have served as the first link in the public safety response chain, having a profound impact on the public safety system in the United States and inspiring replication internationally. Over six decades and billions of 911 calls, 911 personnel have assisted in saving lives and provided invaluable emergency and non-emergency services to local communities. 911 centers’ ability to continue to provide timely support is being challenged by unprecedented staffing shortfalls. The International Academies of Emergency Dispatch reports that some centers are struggling with as much as 30-50% vacancies. With a decrease in the number of personnel, and an increase in the number of annual emergency and non-emergency calls, 911 centers require transformational solutions.
The impact of labor shortages
Between the stressful nature of the job and the fatigue that comes with shift work, the emergency communications profession has always seen turnover. This turnover is becoming more of a challenge though, as the time it takes to onboard a specialist is increasing, and citizen expectations are shifting, creating a need to learn and understand new technologies, such as mobile applications and more.
“The evolution of access to technology and data has made a profession in emergency communications more demanding and has resulted in vastly greater training requirements to perform essential functions in the public safety system. Unfortunately, the existing labor shortage and absence of qualified candidates to fill this essential role has made it difficult to meet industry standards across the nation” said Michael Brewer, Deputy Director of the Jefferson County Communications Center Authority. Recent staffing shortages have forced select 911 centers to fall below the following National Emergency Number Association (NENA) standard: “90% of all 911 calls SHALL be answered within (≤) fifteen (15) seconds. Ninety-five (95%) of all 911 calls SHOULD be answered within (≤) twenty (20) seconds.” Many 911 dispatch centers now have the same dispatcher handling both 911 emergency calls and non-emergency questions, with non-emergency calls taking up the time of highly-qualified and trained dispatchers. With labor shortages, every 911 department across the country is thinking of innovative ways to reduce the call burden on 911 dispatchers and maintaining the same level of customer care to residents.
Automating non-emergency 911 response
Amazon Web Services (AWS) is helping the 911 community mitigate the impact of staffing shortages on call wait times by providing 911 centers the means to automate the receipt and resolution of designated non-emergency calls. This can be done in two ways. First, non-emergency calls placed to administrative 10-digit lines can be addressed with Amazon Connect, leveraging interactive voice response (IVR) to provide automated call resolution and reduce the burden on personnel. Second, non-emergency calls placed to 911 can be transferred with ease to the IVR solution, if the trained professional determines it does not warrant an emergency public safety response. Through the use of IVR, artificial intelligence (AI), and machine learning (ML), AWS can help reduce the burden non-emergency calls placed on 911 centers and allow personnel to focus on life and property saving support.
Amazon Connect is an omnichannel cloud contact center that helps organizations provide customer service at a lower cost. In the emergency communications space, calls can be routed to Amazon Connect through a standard 10-digit number, or through transfer from an emergency call-taker. Amazon Connect is backed by Amazon Lex and Amazon Polly, both of which assist with providing a conversational experience for callers; similar to Amazon Alexa.
Amazon Lex is a fully managed AI service with advanced natural language models to design, build, test, and deploy conversational interfaces in applications. Lex is used to capture the intent of the caller. For example: when a caller asks questions like “What is the phone number of the waste management department?” Lex captures the intent of the caller. Amazon Polly is a service that turns text into life-like speech, to create applications that talk, and build entirely new categories of speech-enabled solutions. Amazon Polly is used to read out the agency prompts and answers to callers’ questions.
The Jefferson County Communications Center Authority is already seeing improvements with this solution, stating that, “On average 10%-12% of daily call volume has been offset by the AWS solution, and this will grow as more data is gathered and expansion of the capabilities are recognized and employed. The longer-term benefit is that it will allow the ECC to get ahead of the retention cycle by providing additional cushion in the immediate staffing needs,” said Brewer.
Another main component in the solution is Amazon Kendra. Amazon Kendra is an intelligent search service powered by ML. Amazon Kendra provides an option to upload questions and answers in familiar formats (e.g. with Microsoft Excel), and when callers ask questions, Amazon Kendra intelligently searches through the provided content to respond back with the respective answer. Through the use of ML, Amazon Kendra enables agencies to search for an answer even when callers only provide part of the question. For example, instead of asking “How can I make my payment?” callers say “Make my payment,” and Amazon Kendra navigates to the appropriate answer.
This solution uses additional AWS services to help connect everything. Amazon Pinpoint, a flexible and scalable outbound and inbound marketing communications service, connects with customers and citizens over channels such as email, short message service (SMS), push, voice, or in-app messaging. Amazon Pinpoint provides an option through which callers can have the information texted back to them during the call flow. When a caller decides to have the information they are seeking texted back, the Amazon Pinpoint service sends a text to the caller’s phone number. AWS Lambda is a serverless, event-driven compute service that lets you run code for virtually any type of application or backend service without provisioning or managing servers. Lambda brings the overall solution together.
Figure 1 illustrates this solution’s architecture:
Figure 1. The architecture for a non-emergency response solution.
Agencies deploying the solution in Figure 1 are reducing their call wait times and are also analyzing non-emergency calls to better respond to their communities.
The solution provides a built-in analytics dashboard through Amazon Kendra to monitor what callers are asking and also identify questions that do not return results. Figure 2 features the analytics dashboard in Amazon Kendra that shows whether and how often Amazon Kendra was able to return a result (pictured here with an example use case). Based on the missed questions, you can decide to include new responses to questions that need to have updated answers included for future inquiries.
Figure 2. The analytics dashboard in Amazon Kendra.Note: 0.00 indicates no results and 100.00 indicates a result was found.
Through the upload feature in Amazon Kendra, customers can upload a file like a Microsoft Excel sheet with new questions and answers (Figure 3).
Figure 3. An example of questions and answers written within a Microsoft Excel spreadsheet, which can be uploaded to Amazon Kendra to include in the non-emergency response workflow.
Conclusion and next steps
A city police department, serving a population of over one million citizens, is also seeing immediate results from the solution. Said one official from the department: “Using Connect, we have been able to significantly reduce citizen hold times and free up our agents to focus on high priority calls. Callers are sent a link for issues such as noise complaints or trespassing on private property reducing their wait time in many cases by up to 45 minutes. In January 2023, we received 55,000 calls with 70% completed in under two minutes. Within two weeks of implementing Connect, our citizens had taken notice, with one citizen saying it was ‘easier to navigate…and an improvement.’”
This solution can assist emergency and non-emergency call takers with their operations and improve community support. Between the automated responses for frequent questions, the ability to analyze questions being asked, and the ability to evolve the responses quickly, emergency communications specialists can spend more time doing what they do best: saving lives.
Read related stories on the AWS Public Sector Blog:
- How Livingston Parish prepares for natural disasters by improving resiliency in the cloud
- How Carbyne’s cloud-native platform helps 911 call centers put eyes on the scene and more
- Modernizing 911 to meet the evolving communications needs of constituents
- Amazon Connect achieves FedRAMP High authorization
Subscribe to the AWS Public Sector Blog newsletter to get the latest in AWS tools, solutions, and innovations from the public sector delivered to your inbox, or contact us.
Please take a few minutes to share insights regarding your experience with the AWS Public Sector Blog in this survey, and we’ll use feedback from the survey to create more content aligned with the preferences of our readers.