Center for Supply Chain Studies

Decision Laboratory

See how a building will operate — before it's built.

These four working models are operational digital twins of a hospital: how patients flow, how rooms get cleaned and turned over, how supplies reach the point of care, and how meals get made and delivered. Each runs on the same shell and reports in the same KPIs the client already manages to.

In production they run on AnyLogic Private Cloud against the project's own Revit geometry — turning a design into an operational forecast while it can still be changed.

Renderings show what a space will look like. These show what it will do.

Four operational twins

Emergency Department

Patient flow & throughput

Arrivals are triaged by acuity, wait for a treatment bay, see a physician, and are discharged or admitted — the classic flow bottleneck.

WatchDoor-to-doc, length of stay, left-without-being-seen
LeverBay capacity & triage staffing — where a surge breaks the department
Open demo →

Environmental Services

Bed turnover & daily cleaning

Discharge bed-turnovers and scheduled daily room cleans compete for one technician pool — and daily work is what slips under pressure.

WatchTurnaround time, daily cleans completed, beds waiting
LeverTechs & travel distance — discharge surges crowd out daily cleaning
Open demo →

Supply Chain

PAR rooms & kanban replenishment

Bins in PAR rooms draw down with care; a room signals at half-par and a runner refills it — a point-of-care stockout if the second bin empties first.

WatchOn-shelf availability, stockout events, replenish time
LeverRunners, par level & travel — PAR placement and building spread
Open demo →

Food Service

Order · kitchen · robot · pickup

Patients order, the kitchen prepares, robots deliver to rooms, and cart staff collect finished trays — three meal rushes cascading through a full day.

WatchOrder-to-delivery, on-time %, trays awaiting pickup
LeverKitchen, robots, pickup & travel — each breaks a different stage
Open demo →

What ties them together

The layout lever

Every twin carries a travel-time control. It stands in for the design itself — adjacencies, floor distances, kitchen and dock placement, elevator access. Stretch it and KPIs degrade in all four; in Food Service one distance slows both delivery and pickup at once. This is how a design decision becomes a measurable, daily operational cost.

One shell, many models

All four share the same controls, the same KPI dashboard, and the same scenario-compare panel. The engine underneath changes; the experience does not. Adding a fifth twin — pharmacy, sterile processing, imaging — is a configuration exercise, not a new application.

How it runs in production

Inputs
Power Apps
Scenario & design parameters from the team.
›
Engine
AnyLogic Private Cloud
Multi-run simulation on the project's Revit geometry.
›
System of record
SQL Server
Every run, input set, and result — versioned and reproducible.
›
Decision view
Power BI
Confidence bands and trade-offs for the client.
Demonstration miniatures. Each runs a simplified model in the browser; production twins run on AnyLogic Private Cloud with real geometry, historical data, and multi-run confidence intervals. Open each twin in its own tab. Center for Supply Chain Studies · Decision Laboratory
0