A while back, a customer was playing around with DTO and noticed something wasn’t working the way they expected. They asked if it was something we could fix. It was.
The Problem
They were trying to model transport operations over a long period of time. Months, not days. Thousands of shifts. Thousands of orders. The kind of scenario that’s easy to describe but hard to solve.
The solver was struggling. First it wouldn’t load. Then it wouldn’t assign any orders. When it finally did, it was painfully slow. Six hours to solve a single scenario.
This wasn’t a minor inconvenience. For clients who want to plan day-to-day operations over months rather than weeks, this was a blocker. The kind of planning that matters most for long-haul linehaul operations simply wasn’t feasible.
What We Did
We went back and looked at how the problem could be broken down.
The key was identifying sub-scenarios within the larger picture. In linehaul optimisation, for example, shifts repeat over a month. The challenge is working out where a long scenario can be split into smaller parts that don’t overlap. Once you do that, you can solve them in parallel.
The user doesn’t need to do anything special. They just put the data in, and the system figures out how to break it up.
The Results
We tested it with two months of data. Over 3,000 shifts. Thousands of orders.
What used to take six hours now takes less than two minutes.
This is coming in the next release of DTO v2. Any client who wants to model day-to-day operations over months, not just days, will be able to use it.
Thanks to the customer who spotted the problem and asked the question.
Why It Matters
Scale has always been the enemy of optimisation. The more variables, the more complex the problem becomes. By finding a way to break large scenarios into smaller, non-overlapping parts, we’ve unlocked a whole new level of planning capability.
For linehaul operators, this means realistic multi-month planning. For everyone else, it means faster solves and more time to act on insights.
