CoDeliver logo

Route Builder & bundling

Route Builder & bundling in practice

A more concrete view of how CoDeliver builds bundled routes: from incoming orders to route proposals and final assignment to drivers.

1. From orders to the Route Builder

The Route Builder works on your actual data: orders arrive from stores, online channels or integrations. Each order carries basic information such as address, preparation time and target delivery window (SLA).

In the dashboard you see, per Zone, which orders are waiting and how they are grouped into proposed routes.

2. How bundling proposals are created

With bundling enabled, the engine tries to combine orders that:

  • are close in time,
  • are reasonably close on the map, and
  • can be served without breaking the SLAs you have configured.

Lateness weights and per‑route limits control how aggressive the bundling is: whether the system prefers more stops per route or faster departures with smaller bundles.

3. Visual routes and candidate drivers

Each proposed route appears as a coloured line on the map, with its stops. At the same time you see which drivers are available, what routes they already have and how much load each one is carrying.

Depending on your plan:

  • in the Semi-Auto plan, a human approves or tweaks proposals before they are sent to drivers,
  • in the Auto plan, assignment is fully automatic based on fairness rules and per‑driver limits, with a full audit trail for every decision.

4. Settings worth experimenting with

For most fleets it is worth experimenting with:

  • Route capacity: how many orders each route can contain, depending on your vertical (for example 2–3 in quick service restaurants, more in courier‑style last‑mile).
  • Waiting time: how long the engine waits before “locking in” a route to see if more compatible orders arrive.
  • Fairness limits: maximum number of active routes or stops per driver so that workload stays balanced.

5. Combine with simulation before changing production

Before applying a new bundling setup to your live fleet, you can run simulations with historical data or synthetic scenarios. This lets you see:

  • how average delivery time changes,
  • how many stops per route you get, and
  • how routes are distributed across drivers.

The goal is not to find a single “perfect” configuration, but a balance that fits your stores and your volume.