A clearer route
to the last mile.
Start with a place. Understand the terrain.
Change the demand, then see the plan—and its limits.
Three synthetic case studies. Original V2.7 engine. Evidence beside every result.
Relief: USGS / Mapzen · Boundaries: Natural EarthMap sources & scope ↗
Choose a case on the map or in the list.
A THREE-MINUTE DEMONSTRATION
Follow one plan. Then test its limits.
Three saved scenes for a stable presentation. Each scene keeps its own original cohort.
01 / CHOOSE YOUR SETTING
One case. A different set of conditions.
Original reference cohorts and saved outputs are preserved.
CUSTOM DEMAND / SAME CONFIGURED RESOURCES
What needs to reach each service site?
Enter doses, with an explicit near/remote split. Stock, staff, transport, carriers and budget stay fixed. Demo products keep their own configured packaging and handling assumptions.
| Service site | Configured product | Required doses | Near doses | Remote doses | Service window |
|---|
The selected profile uses one anonymous simulated household per obligation. Generated identities, grouping and household records are simulated. Doses alone do not identify real children or household visits. No clinical eligibility or due dates are inferred from population data.
Requires computation — no result for these inputs yet.
Local calculation & import workflow
The original Python V2.7 engine runs in a separate browser worker. It computes this demand’s normal reference, selected condition, cancellation-only comparator and supplied audits. Computation stays on this device; there is no hosted compute service. Each plan has an 8-second search limit; a feasible plan can be search-limited.
Use a current desktop browser with WebAssembly and module-worker support. Export the request as a reproducibility record. Imports are fingerprint-checked and re-audited using the supplied components. A 120-second worker timeout makes no infeasibility claim.
Inspect data sources and assumptions ↓BEFORE THE RESULT
Inputs for this plan
Loading the evidence catalog…
This scenario could not be loaded.
Planned service
Precomputed simulation · Synthetic dataLatest delivery arrival is elapsed time from each route’s departure to a session’s arrival, including earlier route activities. It is separate from household travel and the full protection clock.
THE DELIVERY NETWORK
A route to planned service.
A DAY IN THE PLAN
From preparation to return.
02 / COMPARE WITHIN THE SAME COHORT
Conditions and policies tell different stories.
COMPUTED PLAN DIFFERENCES
What the computed plans change
03 / LOOK A LITTLE CLOSER
The details behind this plan.
Village quantities & resource allocation Doses, vials, staff and equipment ↗
Household access & unmet original tasks Original tasks stay in the denominator ↗
Method & evidence Assumptions, search scope and source files ↗
Matching inputs, computed outputs and saved views are included.
04 / SOURCES, ESTIMATES & UNCERTAINTY
Every input has an evidence boundary.
Published external evidence, conditional proxy estimates, scenario assumptions and unresolved local inputs are kept separate. Illustrative calculation remains useful while the formal operational gate is blocked.