“Wait, Why Did That Order Just Go on Two Different Trucks?”
A little while ago, our client Visy flagged something interesting.
They had orders that needed to be picked up together. Same location. Same time. Same vehicle. But then delivered separately to different sites.
On the face of it, that makes perfect sense. You’re at a supplier. You load everything onto one truck. Then you drop each order off where it needs to go.
But here’s where things got tricky.
The Problem No One Had Noticed
Without the right logic in place, our solver would sometimes do one of two things:
- Option one: Send the orders on two different trucks. One goes to Site A. Another goes to Site B. Both start at the same pickup point. That wastes capacity and doubles the trip.
- Option two: Put them on the same truck but treat them as separate loads. The truck picks up Order 1, drops it at the warehouse, comes back, picks up Order 2, and then heads out again. That requires a pointless return trip that burns fuel and driver hours.
Neither was correct. Both created duplicated work.
Visy spotted the gap. And they were right to call it out.
The Fix: Grouped Pickups
So we went back into the DTO and added a new feature: Grouped Pickups.
Now, you can manually impose a simple constraint. These orders must be picked up together on the same vehicle.
Once that constraint is in place, the solver handles the rest. It finds the most efficient way to deliver each order to its destination, without splitting the pickup across trucks and without forcing unnecessary trips back to base.
Same truck. Efficient deliveries. No wasted kilometres.
Why This Matters Beyond Visy
This isn’t just about one client’s edge case. Grouped pickups are useful for any scenario where:
Warehouse constraints mean orders need to leave together
Manual intervention is required to override what the solver might otherwise assume
Fundamentally separate orders happen to start in the same physical location
The feature is now live in the DTO. All our clients benefit from it.
Thanks to Visy for surfacing the need. That’s how good software gets better.
