Getting Ready for the Delivery Rush
How fleets can maximize vehicle and driver capacity, improve routing and prepare for seasonal surges in last-mile demand.
Every holiday season, last-mile fleets face the same challenge: How to deliver more packages without necessarily adding more vehicles, drivers and costs.
The answer isn't simply buying or renting additional trucks. Fleet capacity is the result of several interconnected factors, including vehicle availability, maintenance, driver availability, routing and the ability to respond when conditions change during the day.
Seth Hochhauser, CRO at Autofleet, a subsidiary of Element Fleet, says fleet professionals need to look at those factors collectively. “They all are interrelated,” he says.
“Having really strong routing optimization, scheduling optimization and technology capability is the only way to successfully deliver during these hectic high-demand periods.”
- Seth Hochhauser, CRO, Autofleet
How should fleets think about peak capacity?
Hochhauser says fleet planning should begin with an understanding of expected peak demand and the vehicle capacity required to meet it.
Since vehicles take time to acquire, fleets can't wait until demand arrives to address capacity shortfalls. “You have to order vehicles in advance,” Hochhauser says. “Some companies, for peak seasons will actually incorporate elements of dynamic capacity, where they'll do short-term rentals to supplement their year-round fleet.”
Driver availability is another consideration, but Hochhauser says simply having enough vehicles and drivers isn't enough. Fleets also need to maximize the utilization of the resources they already have.
“Having really strong routing optimization, scheduling optimization and technology capability is the only way to successfully deliver during these hectic high-demand periods,” he says.
Where do fleets commonly go wrong?
One problem is continuing to rely on operational habits that may have worked when volumes were lower. Hochhauser gives the example of assigning drivers to fixed geographic zones because they're familiar with them.
“Some providers like to divvy up zones. They say, ‘This driver always does Zone A, this driver always does Zone B. They're familiar with those zones because they've had that route for the last five years, and we don’t want to mess with it.’ ” he says.
Familiarity may be valuable to drivers, but fixed territories can also prevent fleets from deploying their available capacity as efficiently as possible. “Having artificial constraints on top of already demanding periods only makes things more difficult,” Hochhauser says.
The objective, he argues, should be to distinguish between constraints that are genuinely required by the operation and those that exist primarily because that's how the fleet has always worked.
What should determine the routes?
Not every constraint should be removed. A fleet may have contractual service-level agreements, delivery windows, vehicle limitations or driver qualifications that must be respected.
“You have a certain amount of drivers, you have a certain number of vehicles, and you have the real demands of the fleet,” Hochhauser says.
Those demands can include package locations, delivery windows, package weights and other requirements. The solution is to put the actual operational requirements into the routing and scheduling system and allow it to determine the best available solution.
Why isn't daily route planning enough?
Route optimization isn't simply a matter of drawing the shortest routes each morning. With multiple drivers, vehicles, delivery locations, time windows and vehicle restrictions, the number of possible combinations quickly becomes overwhelming.
Even the best plan can become obsolete minutes after drivers leave the yard. A driver could call in sick, or have to leave his shift early for personal reasons. Factor in traffic, accidents and other operational disruptions, and even the best-planned schedules can fall apart.
That means fleets need the ability to re-optimize during the day, rather than simply relying on the morning's plan. And while no human can handle all the moving parts on their own, an effective routing system can quickly work out all the details in real time.
What data does the routing system need?
Hochhauser breaks the required information into several categories. The first is demand: What needs to be delivered? Where does it need to go? When does it need to arrive?
The second is supply, which includes both vehicles and drivers. Vehicle data goes beyond the number of trucks in the fleet. A routing system needs to know which vehicles are actually available and what each one can carry.
“You might have five different vehicles. Not every vehicle holds the same amount or type of cargo,” Hochhauser says. Driver schedules and qualifications also make a difference.
A third layer is infrastructure, including traffic conditions and, increasingly, EV-specific information such as range, battery capacity and charging requirements. Weather can add another variable by affecting traffic and operating conditions.
Where does maintenance fit?
Vehicle availability is fundamental to delivery capacity. A fleet that planned for 100 available vehicles, but has five in the shop during peak season, has less capacity than expected.
“Proper fleet management and proper maintenance is critical to ensuring you have the relevant supply,” Hochhauser says.
Connected-vehicle data can help fleets go beyond fixed maintenance intervals by identifying warning codes or other emerging problems.
That information can also feed directly into capacity planning. If a vehicle becomes unavailable, the system can identify the resulting capacity gap and potentially allow the fleet to arrange a replacement vehicle.
What role does AI play?
AI isn't entirely new to route optimization. Machine learning and data science have long been used to solve complex routing problems.
Hochhauser sees a growing role for AI in making operational data easier for fleet managers to use.
Instead of requiring managers to search through multiple reports and dashboards, an AI assistant can provide a natural-language interface to operational data.
“You're able to actually ask it, ‘Which of my orders are running late? How many vehicles do I have available for tomorrow?’ ” Hochhauser says.
What should fleet managers do now?
For a fleet expecting a significant increase in parcel volume over the next three months, Hochhauser recommends starting with the fundamentals: Understand existing capacity, optimize the resources already available, and secure supplemental capacity if necessary.
A fleet shouldn't automatically assume that more demand means more trucks.“You might have 500 vehicles that you think are capable of delivering 20 packages daily, but they might actually be able to do 30,” he says.
That additional capacity could come from better utilization, rather than additional assets.
Once existing capacity has been optimized, fleets can look at supplemental vehicles. If you think you’ll need extra trucks, Hochhauser recommends negotiating short-term leases sooner rather than later. “Make sure you lock down that capacity before they're all booked,” he says.
For fleet managers, the broader lesson is that peak-season preparation isn't simply about adding vehicles. It is about understanding exactly how much capacity the existing fleet can provide, ensuring those vehicles remain available, and using routing, scheduling and real-time data to get as much productive work as possible from every available vehicle.
