Project Pineapples

Inspiration

The inspiration for this project actually came from something that some fellow classmates built for a class project, which was a vending machine selling hawker food, increasing accessibility to hearty hawker fare 24/7. We decided to expand on this idea, not only for diners but also for hawkers. When a nearby hawker centre closes, hawkers cannot see where displaced demand will go. A closed hawker centre can mean a wasted trip, a longer journey, or fewer affordable choices—especially for elderly and lower-income residents. For hawkers, under-preparing causes stock-outs and lost sales, while over-preparing creates food waste and unnecessary cost. This shared problem inspired the creation of HawkerPulse AI to turn closure information into actions that reduce wasted journeys, missed sales, and surplus food.

What it does

HawkerPulse AI uses a hawker closure information to produce a diner recommendation and a hawker preparation decision, demonstrating a two-sided value exchange.

For Diners (Verified Meal Assistant): The system recommends open, affordable alternatives and provides closure-aware routing, price and source freshness, and honest crowd estimates. For example, if Teck Ghee Court closes, it advises that Teck Ghee Square is approximately 0.8 km away and has three listed meals below $4.
For Vendors (Vendor Prep Planner): It offers advance notice of demand shifts so hawkers experience fewer stock-outs and lower surplus food waste. It calculates demand pressure using top drivers like nearby closures, lunch periods, and rain risk to recommend preparing moderately above usual (e.g., +10-15% versus the vendor baseline). By not giving a specific number, it allows hawkers to adjust their preparation based on their own demand baseline, helping to prevent food wastage while ensuring there is enough food.