AI chatbot ‘SafeRBot’ to revolutionize US emergency call centers

The University of Illinois Urbana-Champaign has unveiled its groundbreaking AI chatbot, ‘SafeRBot,’ designed to assist emergency call centers across the United States. Developed by a team led by Professor Yun Huang, the chatbot is poised to enhance how emergency calls are handled, providing empathetic and structured communication while gathering critical details about incidents.

SafeRBot’s primary function is to process emergency calls by asking essential questions about the location, nature of the incident, the individuals involved, and other key details. What sets it apart is its ability to automatically ask follow-up questions until all necessary information is gathered, ensuring that no critical detail is overlooked. This system not only improves accuracy but also streamlines the reporting process.

Using a cutting-edge large language model (LLM), the chatbot is capable of turning unstructured conversations into structured data for incident reports. It supports both English and non-English speakers, seamlessly switching between languages based on the caller’s input. For instance, if the initial response is in Spanish, the chatbot immediately adapts, conducting the entire conversation in Spanish to ensure effective communication.

Professor Huang emphasized the practical benefits of SafeRBot, highlighting its potential to reduce the workload of emergency dispatchers. “By automating the process of gathering information, SafeRBot minimizes the time needed to document incidents and significantly enhances the quality of reports,” she explained. The system is designed to alleviate the pressure on human operators, reducing the risk of burnout among emergency call center staff.

The chatbot operates through an intuitive interface. Users input incident details on their computer or smartphone, while the chatbot’s questions appear on one side of the screen. As answers are provided, they are automatically organized into the appropriate fields on the incident report form displayed on the other side. This method ensures a streamlined and user-friendly reporting process.

Beyond its technical capabilities, SafeRBot is also programmed to offer varying levels of emotional support based on the needs of the individual reporting the incident. Professor Huang noted that users often have different emotional requirements when describing distressing situations. “SafeRBot allows reporters to personalize their experience, offering empathetic support where needed,” she said. Research has shown that empathetic engagement increases users’ willingness to provide detailed answers, making the chatbot not only a tool for efficiency but also for human-centered interaction.

The SafeRBot initiative represents a significant step forward in leveraging artificial intelligence to improve public safety infrastructure. By offering a multilingual, empathetic, and efficient solution, it has the potential to transform emergency call handling, ensuring faster response times, better data accuracy, and reduced strain on human dispatchers.

Photo: Dr. Yun Huang, Associate Professor, iSchool @ UIUC