Routite Optimize · GPU route solver

Re-solve the whole region, not just one route

Optimize is the engine. It uses NVIDIA cuOpt on GPUs to re-solve routing, dispatch and scheduling with every constraint you run on, starting from the plan drivers are already following.

Inside a re-solve

  1. InputLive state from Twin
  2. Warm startCurrent plan
  3. SolvecuOpt on GPU
  4. OutputRoutes, diff, reasons

Re-solve summary · Zone 4, 10:47

Example
  • EventVan 3 down
  • Stops moved11
  • Routes changed2 of 5
  • Solve timewarm start6.2 s
  • Windows at risk2 late
Distance before
184 km
Distance after
171 km
Late stops before
11
Late stops after
2
Example Values are illustrative and show the shape of an Optimize result, not a measured run.

Stop 01 · Capabilities

What Optimize does

Engine

GPU vehicle routing

cuOpt solves capacitated, time-windowed routing with pickups, deliveries and multiple depots.

Stability

Warm starts

Each re-solve starts from the current plan and penalises needless changes.

Scope

Whole-region solves

Re-plan the full network, not one driver at a time, so stops can move between vans.

Objectives

Your trade-offs

Weight distance, lateness, overtime and emissions; change weights per region.

Explain

Reasons with every plan

Each moved stop carries why it moved, ready for Copilot to explain.

Scenarios

What-if solves

Run the same day with one more van or a different cut-off and compare.

Stop 02 · Constraint catalogue

The rules Optimize respects

Filter by kind of rule and by release. Anything missing for your fleet is the first thing to raise in a pilot.

Kind
Status
Constraint catalogue. Core means planned for the first release; planned means later.
ConstraintWhat it meansKindRelease
Time windowsHard or soft windows per stop, with lateness penalties.TimeCore
Vehicle capacityWeight, volume, pallets or any count, per compartment.LoadCore
Driver hoursShift length, breaks and maximum drive time.TimeCore
SkillsJobs that need a certified technician or a two-person crew.PeopleCore
Vehicle typeStops that need a tail-lift, a small van or a reefer.LoadCore
Pickup and deliveryPaired stops in the right order on the same vehicle.LoadCore
Multi-depotVehicles that start and end at different depots.NetworkCore
ZonesKeep drivers inside familiar areas unless it pays to cross.NetworkPlanned
Priority ordersOrders that must be served even if others drop.TimeCore
TemperatureChilled and frozen compartments with door-open limits.LoadPlanned
Change penaltyPrefer plans that leave on-track drivers alone.PeopleCore
EmissionsWeight CO₂ per km by vehicle and add it to the objective.NetworkPlanned

Stop 03 · How it works

How Optimize works

  1. 06:00full solve

    Morning plan

    All orders, vehicles and constraints are solved before the shift.

    Input Orders, fleetOutput Day plan
  2. on eventwarm start

    Re-solve

    Twin hands over the changed state; Optimize re-solves from the current plan.

    Target <10 s regional
  3. afterdiff

    Explain the change

    Moved stops, new ETAs and reasons are passed to Copilot for review.

    Output Diff and reasons

Stop 04 · Specification

Specification

Draft specification. Figures are design targets, not measured results.
ItemDetailStatus
SolverNVIDIA cuOptIn build
Problem typesVRP with time windows, capacities, pickup and delivery, multi-depotIn build
Regional re-solveAbout 1,000 stops in under 10 sDesign target
Full-day solveUp to 10,000 stops per regionDesign target
ObjectivesDistance, lateness, overtime, emissionsIn build
DeploymentRoutite cloud; VPC and on-premisesPlanned

Stop 05 · Place in the loop

Where Optimize sits

Optimize takes the live state from Twin and hands its plan to Copilot for review. It is the step every other product depends on.

Return leg Edge sends real arrival and service times back to Twin. ETA and demand models retrain on them, so tomorrow's first plan starts closer to reality.

Stop 06 · Questions

Questions dispatch teams ask

Is Optimize a replacement for our route planner?

It can be. Many teams will start by running it beside the current planner during a pilot and comparing.

What if our constraint is not in the catalogue?

Tell us. Encoding unusual business rules is part of every pilot.

Do we need our own GPUs?

No. Routite cloud runs them. On-premises deployment on your own GPUs is planned.

Stop 07 · Destination

Next step

See Optimize on your own data

A replay of one past day shows what this product would have done, with every assumption written down.