Inspiration
Hawker centres are where Singapore eats, meets and catches up. But when we looked at the open data, we found no single view that answers three planning questions together:
- Which neighbourhoods have poor access to affordable hawker food?
- Who is left without a nearby alternative when a centre closes?
- Which centres could realistically start separating food waste?
The numbers made this feel urgent. NEA's 2026 schedule lists 123 markets and hawker centres with quarterly cleaning closures. 38% of them also have recorded repairs or redecoration works, so closures are not just routine cleaning. Meanwhile, only 18% of the 784,000 t of food waste generated in 2024 was recycled.
What it does
KopilamAI is built for NEA and town-council planners. It:
- Scores walkable access to hawker food per neighbourhood, weighted by the stalls within walking distance rather than the number of centres per planning area, and shows where the two measures disagree.
- Simulates closures. Pick a centre, and the app flags neighbourhoods left with no walkable alternative and counts the residents affected.
- Ranks centres by food-waste separation readiness, with waste shown as a range (conservative / base / ambitious) from a published scenario framework, not as a single false-precision number.
- Explains findings in plain English. AI briefings and a Genie space quote only numbers computed in the pipeline.
Built With
- apache-spark
- data.gov.sg
- databricks
- databricks-ai/bi-dashboards
- databricks-genie
- databricks-sql
- delta-lake
- foundation-model-apis
- genai
- geojson
- geospatial-analysis
- lakeflow
- mlflow
- pyspark
- python
- unity-catalog
Log in or sign up for Devpost to join the conversation.