This post is also available in:
עברית (Hebrew)
Emergency dispatch centers often grapple with staffing shortages and increasing workloads. Now, a new AI-driven solution aims to ease the pressure on human operators while improving the reporting experience for the public. SafeRBot, a chatbot developed by researchers at the University of Illinois Urbana-Champaign, leverages advanced AI technology to help report non-emergency incidents quickly and effectively, all while offering empathetic support.
According to TechXplore, SafeRBot uses a large language model (LLM) to guide users through the process of reporting non-emergency situations, transforming unstructured conversations into structured reports. The bot asks a series of consistent, informative, and empathetic questions, ensuring that all the necessary details are collected while also providing emotional support to users who may need it. SafeRBot is designed to work in both English and multiple other languages, automatically switching languages based on user input.
The researchers explain that SafeRBot aims to assist dispatch centers by providing a reliable alternative when human dispatchers are unavailable or overwhelmed. By automating follow-up questions and gathering incident details more efficiently, SafeRBot helps reduce dispatcher workload, ultimately preventing burnout and improving the quality of the information collected.
The system works through a simple interface: community members visit the SafeRBot website, where they begin answering questions on the left side of the screen. The bot’s responses and follow-up questions are reflected on the right side, automatically filling in fields of an incident report. SafeRBot’s multilingual capability ensures inclusivity, allowing users to continue the conversation in their preferred language, such as Spanish, if needed.
One of SafeRBot’s most innovative features is its ability to adjust the level of emotional support it provides, according to TechXplore. The team’s research reveals that different users require varying levels of emotional support when reporting incidents. SafeRBot can be tailored to offer compassionate responses, ensuring that users feel heard and supported during the reporting process.
SafeRBot is expected to significantly ease the strain on dispatch centers. Once fully deployed, police agencies will be able to access incident reports directly through a secure, encrypted dashboard, facilitating smoother integration with existing systems. The data is stored safely on Amazon Cloud, with multiple layers of security to protect user privacy.
With real-world testing and collaboration from the Urbana Police Department, SafeRBot is poised to make a meaningful impact on community safety reporting, improving both the efficiency and emotional well-being of users.
The team’s paper was published on arXiv.
