After hurricanes, utility companies face numerous unexpected challenges. Their usual rescue plans often fall short because each disaster brings unique problems. Residents in the affected areas might not get the specific help they need quickly using traditional methods.
GPTProX is a solution to this. It combines ChatGPT’s knowledge with utility companies’ specific data. When a resident asks for help, GPTProX consults databases like the utility asset database, GIS, and ontology databases to give accurate and location-specific advice.
ChatGPT suggests possible responses based on the resident’s query. These suggestions are then checked against the Prolog Knowledge Base to ensure the advice is relevant. If there are any doubts about the advice, the system uses Category theory to clear up any inconsistencies.
The system doesn’t just give general information. It offers clear, actionable steps. For example, it can guide residents to safer areas, tell them how to access emergency supplies, or help them report specific damages for faster repairs.
The best feature of GPTProX is its Multi-Agent Optimization Solver. After a hurricane, it gives the most effective advice by considering various factors to optimize the help provided to the residents.