Why traditional AI fails fleets
Jake Dettmer, Optimal Dynamics SVP of product, told FleetOwner that most current AI "agents" are essentially chatbots following simple rules without understanding logic or math.
"An agent—if built to automate a task, and it’s the wrong task—will drive profitability to the floor at a faster rate than a human could," he warned.
To prevent this, Scale uses stochastic optimization. Instead of looking at a single load in a vacuum, it calculates the "marginal profitability" of a move—factoring in where that driver needs to be five days from now to keep the truck profitable.
During an earlier press conference at Truckload 2026, OD Chief Executive Daniel Powell illustrated this with a common scenario: A shipper emails a load tender from New York to Chicago at a solid rate. A standard AI agent sees the high rate per mile and accepts it. However, the AI might fail to realize that the driver will have to “deadhead” 500 miles back home afterward, turning a profitable load into a net loss.
The Optimal Dynamics’ Scale solution
Scale is designed to handle that first—and most critical—decision a carrier makes: which freight to accept. Here is how it makes those judgments:
- Contextual intelligence: The system proactively detects network imbalances days in advance and identifies the specific loads needed to bring the network back to neutral.
- Autonomous negotiations: Scale deploys agents that can actually negotiate rates with shippers or change appointments on behalf of the carrier based on the "marginal profitability" of the move.
- Procurement power: If a carrier hits its commitment with a customer, Scale automatically scours load boards and broker APIs to find the right backhaul at the right rate.
The trust factor
Dettmer acknowledged that handing over these decisions requires deep trust. Optimal Dynamics’ engine is built on 40 years of research out of Princeton University and uses "stochastic optimization" to prioritize long-term profitability over historical data.
"The engine will make decisions that are much different than the heuristics that drive how a human makes the decisions," Dettmer told FleetOwner. "But that computing power doesn’t exist in one individual."
He said the system also uses “shadow planners and shadow dispatchers” to ensure the system’s proof of value in the process and that Scale is optimizing the right operations and freight networks.
Scale aims to shift the CSR role from manual data entry to strategic oversight. As the industry emerges from a sluggish freight economy with more automated technology options, Optimal Dynamics believes the winners will be fleets that find the right balance between human intuition and machine-driven math.
