Page 286 - AI for Good Innovate for Impact
P. 286

AI for Good Innovate for Impact



                      By integrating Natural Language Processing (NLP) with geospatial data and real-time datasets,
                      the chatbot enables:

                      •    Guided evacuation based on current conditions
                      •    Locating nearby shelters and essentials
                      •    Crisis communication and mental health support
                      •    Missing persons reporting and follow-up
                      •    Automated alerts and verified updates

                      The chatbot is accessible via WhatsApp, SMS, Telegram, and voice, ensuring broad usability.
                      The system also supports responder dashboards through API integration and aims to improve
                      situational awareness using AI.

                      Use Case Status: Ongoing

                      Category: AI for Disaster Response & Climate Resilience

                      Partners: Ethiopian Red Cross, Ethiopian Meteorological Institute

                      2�2     Benefits of use case

                      •    By guiding people to medical assistance and providing psychological support
                      •    Strengthens emergency response systems through AI-driven automation.
                      •    Ensures better preparedness and quick response to urban disasters.
                      •    Uses AI to predict, detect, and mitigate the impact of climate-related disasters.

                      2�3     Future Work

                      •    Expansion of local language support and NLP accuracy
                      •    Data collaboration with National Meteorological Agency and Geospatial Institute
                      •    Deployment pilots with humanitarian agencies
                      •    Anonymized dataset publication through GitHub or official repositories
                      •    AI model fine-tuning with disaster response specialists
                      •    Open-source roadmap and interface standardization
                      •    Integration with international alert systems (e.g., GDACS, UN OCHA)


                      3      Use Case Requirements

                      •    REQ-01: It is critical to develop NLP models for local languages.
                      •    REQ-02: It is expected to integrate with real-time shelter and evacuation data.
                      •    REQ-03: It is critical to ensure chatbot access through multiple platforms (SMS, Telegram,
                           WhatsApp).
                      •    REQ-04: It is of added value to include voice interface for accessibility.
                      •    REQ-05: It is expected to anonymize and geotag data for responder dashboards.
                      •    REQ-06: It is of added value to offer psychological support content during emergencies.













                  250
   281   282   283   284   285   286   287   288   289   290   291