Automate and optimize facility location
Every network planner has a list of candidate sites, a demand forecast, and a budget. Few have anything that decides which sites to open and who each one serves, so that no site is overloaded and the network isn't paying for sites or miles it doesn't need. Let's solve that.
One solve picks the sites and assigns every consumer
Timefold decides which facilities to open and which consumers each facility serves as one optimization problem. Facility capacity is enforced as a hard constraint, so every plan can actually serve its demand. Within that rule, the model weighs the setup cost of every opened site against the distance between consumers and the facility serving them.
Where Timefold goes beyond spreadsheet network design
Four things decide whether a network holds up. Timefold bakes them into the model.
Capacity as a hard rule
No facility serves more demand than it can handle, and any shortfall shows up as an explained violation.
Cost and distance in one score
Setup cost and travel distance are weighed together instead of in separate spreadsheets.
Real distances, not straight lines
Distances come from the map service, so the plan reflects how far consumers actually are from each site.
Scenarios in minutes
Change a capacity, a cost, or a candidate site and solve again to compare networks side by side.
Where you build decides what you pay for years
Opening a warehouse, depot, or store is a long-term commitment. Choose the wrong sites and every delivery travels further, every year, for as long as the network stands.
A site decision stays in place
Leases, buildings, and staff make locations hard to change later.
Pays for the distance to its facility
Small differences in location add up across thousands of shipments.
Pull in opposite directions
More sites cut travel distance, but each one adds setup and fixed costs.
Ways to pick sites from 50 candidates
Too many combinations to compare by hand or in a spreadsheet.
Constraints the model expresses today
Capacity is never exceeded
The total demand of the consumers a facility serves stays within that facility's capacity. Any overload is a hard violation, sized by the excess demand.
Every consumer gets a facility
Each consumer is served by exactly one facility, so no demand point is left out of the plan.
Pay only for facilities you open
A facility's setup cost counts as soon as it serves at least one consumer, so the model only opens sites that earn their keep.
Consumers close to their facility
The model minimizes the total distance between every consumer and the facility serving it, using distances from the map service rather than straight lines.
Setup cost and distance in one trade-off
Setup cost and distance are weighted against each other, so the model weighs a cheaper site against the extra distance it creates for consumers.
Every penalty explained
Each overloaded facility, opened site, and consumer distance comes with a justification, and the plan reports facilities used, total setup cost, and total distance.
FAQ
Do you have questions? We have answers.
Picking the closest site for one consumer is easy. Picking which sites to open at all, while each one stays within capacity and the setup cost of every opened site pays for itself, is not. Open one more site and the best assignment changes for every consumer nearby, and so does the load on every other site. That is a constraint problem, not a map lookup.
Capacity comes first and is a hard constraint. Within that, the model minimizes two soft goals together: the setup cost of every facility that serves at least one consumer, and the total distance in meters between consumers and the facility serving them. Distance carries a weight of five and setup cost a weight of two, so you can see and tune how much extra distance one avoided site is worth.
Every consumer still gets a facility, so the plan stays complete, but the score shows a hard violation for each overloaded facility, sized by the excess demand. The justification names the facility and its demand, so planners see exactly where capacity runs short and how much to add.
The Timefold Platform map service builds a distance matrix for every facility and consumer location before each solve. The model reads distances from that matrix, and locations the map service can't resolve are reported back instead of being silently dropped.
No. Timefold sits as an optimization layer between the systems that hold your candidate sites, costs, capacities, and demand, such as an ERP, a data warehouse, or a GIS, and the people who make the site decision. It reads candidate facilities and consumers and returns which facility serves each consumer. It doesn't replace your GIS, your real estate process, or your financial model.
Any problem where consumers with a demand are served by one of several candidate facilities, each with a capacity and a setup cost. That covers distribution centers, stores, test centers, and telecom masts, among others.
When scheduling works, everything works.
Less waste. More control. Teams that trust the plan.