Skip to content

Talking Decision Analytics Series Part 6: Balancing Objectives for Real-Time Dispatch

Dispatching drivers is a balancing act that can feel like being pulled in different directions at the same time: reducing empties, maximizing on-time pickup, getting drivers home, improving driver pay, and making sure you cover your major accounts.  

We start with the goal of maximizing total revenue minus the cost of moving the loads (including empties). Most of the time a carrier has to pay for empties, but an owner-operator may be responsible for his own empties, creating the temptation to assume that empties are free. It is very important for the model to recognize that empties are not free - someone has to pay for those miles, so we have to tell a model what empties actually cost, even if it is the driver paying for them.

We then introduce the other goals (on time pickup and delivery, getting drivers home, …) through a system of bonuses and penalties. We might charge a $200 penalty if we do not get a driver home on time, or a $50 penalty for picking a load up load. Perhaps we want a target of at least 1800 miles per week for each driver, so we include a penalty if a driver’s rolling mileage falls below 1600 miles for the last 7 days.

Each carrier may wish to tune their bonuses and penalties differently. One carrier may feel that minimizing driver turnover is the foundation of a well-run company. Others may wish to focus on high-quality service. We tune these parameters by running simulations and then reporting on time-at-home, on-time pickups and deliveries, as well as net operating revenue per loaded mile (to name a few). These parameters can be adjusted to reflect market conditions: a strong market requires holding onto drivers, while a weak market needs a focus on customers.  

A green semi-truck driving down a highway.