Researchers at the Pennsylvania College of Information Science and Technology have introduced an AI-powered emergency call tool designed to speed up critical care for pregnant women in Kenya. The project, highlighted on the university’s site, demonstrates how intelligent systems can shorten response times in life-or-death situations, potentially saving lives through faster triage and outreach.
In Kenya, a significant share of maternal and infant mortality stems from delays in receiving timely medical attention. The system, named TRIage, analyzes incoming messages from expectant mothers using artificial intelligence. When a message signals danger or urgent needs, it is automatically routed to a dedicated, small response team that can mobilize help right away. This workflow aims to cut through communication bottlenecks and connect women with practical support as swiftly as possible.
Initial results from field use show meaningful improvements in response times. On average, the time from a girl or woman requesting help to a on-ground or remote assistance being delivered has dropped to within about an hour. The program reports that roughly 85% of the identified cases required hospitalization, a statistic that underscores the system’s effectiveness in identifying patients with significant medical needs and ensuring they receive appropriate care promptly.
Today TRIage serves a large and growing user base in Kenya, with more than 350,000 pregnant women relying on the service. The platform processes around a million messages each month, a volume that highlights ongoing demand for faster, scalable support in maternal health. The scale also reveals a challenge: the volume of messages far exceeds what the small team could handle without automation, making the AI-assisted approach essential to sustain timely help. Researchers emphasize that this technology complements existing healthcare infrastructure and telephone-based support, filling gaps where traditional systems struggle to keep pace with demand. The overarching goal is to reduce preventable delays, improve triage accuracy, and empower healthcare workers to act quickly when a crisis signals itself through a simple message. The project continues to explore refinements in signal detection, language processing, and user experience to ensure that the system remains accessible, reliable, and inclusive for women in diverse communities. (Source: University researchers and project updates)