The FreightMath Optimizer evaluates billions of possible freight combinations to identify the connected loads that can produce the highest network margin under realistic deadhead, capacity, margin, and customer-commitment rules.
Density, efficiency, and velocity are essential to trucking profitability, but daily operations make it difficult to evaluate every load. The optimizer was developed over more than three years to separate the margin winners from the weak freight in a carrier's existing or prospective network.
The Questions the Optimizer Addresses
KSMTA clients regularly asked:
- How can the carrier maximize margin from current customers and lanes?
- How many trucks should the network support?
- Which logistics loads should move on company assets?
- How will newly awarded lanes affect the network?
- What happens after losing a customer?
The optimizer answers those questions using a modified traveling-salesman model, a form of combinatorial optimization.
Building the Freight Basket
The process begins with a defined basket of freight for a selected period. For FreightMath clients, the recurring analysis uses the previous four weeks of loads actually hauled.
For additional decisions, KSMTA can add freight sold by the logistics division, expected or proposed shipper awards, or future load opportunities. Lost freight can be removed to model the resulting network.
Applying Practical Constraints
The model's sole mathematical objective is maximum margin. Constraints make its result operationally realistic.
Maximum Deadhead
Because the model creates its own empty moves to connect profitable loads, it could otherwise produce impractical repositioning. A mileage limit restricts those empty movements.
Minimum Margin Threshold
FreightMath focuses on direct and variable cost. A network can appear profitable on that basis while still failing to cover fixed and back-office expense.
The Minimum Margin Threshold requires enough gross margin to support the company's broader commitments. KSMTA recommended a threshold between the current network gross-margin percentage plus 3% and plus 10%, depending on the realistic potential of the carrier's OTR network.
The threshold can also be raised over time to drive continuous improvement.
Capacity
When logistics or prospective freight is included, the optimized network could exceed the actual fleet. Conversely, a carrier may want to know which freight to retain while reducing trucks.
A total-mileage constraint limits accepted loads to available capacity while preserving the most profitable connected freight.
Freight Commitments
The optimizer uses a binary decision for each trip: accepted or rejected. Some customer commitments must remain, so the model can designate them as locked trips.
The article cautions against overusing this rule; a business should not treat any customer or lane as a permanent golden cow.
The Output
The objective is to maximize network profit while ensuring that each accepted trip connects to another accepted trip.
A companion dashboard compares loads actually hauled with loads the optimizer would accept.
Customers and Lanes
Users can drill from a customer into its lanes or from a lane into its customers, isolating freight that requires a price increase or a capacity shift.
Capacity
Using the Minimum Margin Threshold and average dispatched miles per tractor per week, the dashboard calculates the tractor count needed to haul the selected freight and achieve the target margin.
Automatic Tour Identification
The next planned capability was automatic identification of repeatable freight tours—consistent load patterns that can be connected for more efficient capacity assignment.
The FreightMath Optimizer was designed not merely to improve routes, but to help carriers decide which freight belongs in the network and how much capacity that profitable basket can support.