CHOOSE AN EXPERIMENT
This reproducible five-day study is ready to explore immediately.
ROUTE SIMULATOR
Replay 5 illustrative service days. At each decision time, BakedBoston checks available routes, coordinates drivers, and recommends practical pickups and deliveries.
INTERACTIVE REPLAY
ROUTES AT 1:24 PM
BakedBoston coordinates the requests together so the same bakery pickup is not offered twice.
DRIVER REQUEST
These preferences help rank options. A slightly different route may still be recommended when it offers more overall community benefit.
ROUTE MAP
Dashed lines show alternatives. The solid orange line is the recommended route.
ROUTES RECOMMENDED TO DRIVER 2
These are alternative options for one delivery. Only one route is chosen.
DRIVER SELECTION
This route is the best overall available match for this driver, bakery, and pantry. Other routes shown are alternatives, not rejected trips.
MODEL COMPARISON
Every approach sees the same 5-day scenario.
Each approach puts more emphasis on a different goal, such as pantry coverage, environmental impact, fair distribution, or volunteer choice. The results show those trade-offs side by side.
SIX NON-DUPLICATED OUTCOME PILLARS
A post-hoc communication index that gives equal weight to six non-duplicated outcome pillars. Each pillar is scored from 0 to 100 relative to the best policy in this exact scenario.
A simple snapshot across service, food, environment, fairness, volunteer fit, and efficiency. The full results below show the trade-offs behind it.
Maximizes one acceptance-adjusted expected-impact objective across simultaneous drivers, balancing food, fairness, environment, pantry priority, and driver fit.
A minimal volunteer-route adaptation of the 2018 periodic unpaired pickup-and-delivery model: protect distinct pantry service first, then completed pickups and miles.
Read the source study ↗A minimal volunteer-route adaptation of the 2025 Total-Curb model: protect service first, then minimize total direct system CO2e.
Read the source study ↗A minimal adaptation of the 2021 SLSF-noZ model: optimize short personalized menus over seeded stochastic willingness scenarios, then assign among routes drivers are willing to fulfill.
Read the source study ↗A deliberately limited adaptation of the 2018 food-rescue allocation work: protect completed pickup service, then minimize the maximum gap in cumulative ultimately distributed food across pantries.
Read the source study ↗Each measure is shown separately so you can see what every approach gained or gave up.
Each approach is tested on the same simulated day.
How much food is recovered for people and how much is left over.
Environmental results include travel and food that is not ultimately used.
How convenient the selected routes are for volunteers.
Every model sees the same simulated day, so the results show how different routing priorities create different outcomes.
Use the detailed metrics to explore the trade-offs. The source-study links describe each comparison approach in depth.
This saved example is an illustrative scenario, not proof of average performance.
HOW THE SIMULATOR WORKS
The simulator creates a shared, synthetic food-rescue day and asks each approach to make route decisions. BakedBoston checks whether a route can happen, then weighs food recovery, pantry service, environmental impact, and volunteer practicality.
The comparison is designed to show trade-offs, not to declare a single universal winner. It does not contact real partners or issue delivery assignments.
A full technical paper will be linked here when it is complete.
RESEARCH FOUNDATION
These studies inform the ideas represented in the simulator. Their links provide the underlying technical detail.