The Problem.
Shared laundry facilities in residential buildings suffer from predictable demand spikes — Sunday evenings, Saturday mornings — that leave machines unavailable for hours while they sit idle at other times. The solution isn't more machines; it's understanding and redistributing the demand.
Visualization
Usage Heat-Maps
Algorithm
Linear Demand Sorting
Data Source
Historical Resident Usage
Output
Optimal Time Suggestions
Usage Heat-Map.
Machine Contention — Simulated Weekly Usage Pattern
Mon
Tue
Wed
Thu
Fri
Sat
Sun
8am
10am
12pm
6pm
8pm
10pm
Darker blue = higher contention. Sunday 6–8pm and Saturday 10am–12pm are peak congestion windows — primary targets for demand smoothing.
How It Works.
01
Historical Data Ingestion
Collects timestamped machine usage records from the facility over weeks to months. Builds a representative picture of demand patterns across time-of-day, day-of-week, and seasonal variation.
02
Heat-Map Visualization
Renders the demand surface as a color-encoded grid — immediately surfacing which time slots are oversubscribed and which have spare capacity. Makes the optimization problem visible before solving it.
03
Linear Demand Sorting
Backend algorithm prioritizes machine scheduling by sorting resident requests against predicted availability windows. Shifts peak demand toward low-contention slots by surfacing real-time and predictive availability to users at the point of decision.
Constraint → Insight.
The interesting part of this project wasn't the algorithm — linear sorting is simple. It was recognizing that the constraint is behavioral, not mechanical. Residents cluster into the same time slots by default because they have no visibility into demand. The heat-map is the intervention: give people the data and the decision changes itself.