PS-19Livelihoods & Clean Energy
GridSathi
AI Energy Forecasting and Smart Demand Scheduling
SDG 7SDG 9SDG 13
The challenge
Campuses, hostels, small facilities and microgrids may experience avoidable peak demand because major loads are scheduled without considering predicted consumption, renewable generation or tariff periods. Develop an AI-assisted energy management system that forecasts short-term demand and, where relevant, solar generation, then recommends a practical schedule for flexible loads. The system should show expected savings or peak reduction and maintain user/administrator control instead of switching critical loads autonomously.
What a working submission needs
Create a dashboard prototype using historical or synthetic load and solar data.
It should forecast demand, identify peak periods, recommend rescheduling for at least three flexible load types, and compare baseline versus proposed energy/cost/carbon metrics.
Stretch goals
battery scheduling, anomaly detection, occupancy-aware forecasting or live IoT integration.
Optional. A working core counts for more than feature count.Suggested stack
Web/mobile application, AI/ML models, a mapping or dashboard interface, and open APIs as appropriate to the problem.
Scope and responsibility
Healthcare, safety and public-sector challenges are decision-support tools, not autonomous clinical, legal, emergency-command or enforcement systems. Name your prototype’s limitations in the demo. Judging weighs a working core, justified use of AI, technical quality, real-world impact and responsible design above feature count.
Also in Livelihoods & Clean Energy