Route Planning for LTL Operations Using AI Assignment

Ron Dombrowski January 22, 2020

Less than truckload (LTL) carriers provide both first mile and last mile services for organizations that move goods.

Here is the best definition that I found:

The first and last mile of product distribution are very crucial to any shipper. The first mile refers to the movement of products from a shipper to a courier service or to anyone who will take these goods to their final users; and the last mile refers to this final movement of products to their final users.

Truck on bridge

LTL carriers provide services to both pickup goods from a shipper’s location and then, after moving them through their own network, deliver those goods to their final destination.

Since the same trucks are used to do both the pickups and deliveries, there is a large opportunity to do reduce costs by doing this in an optimal manner.

Let’s take a look at the process from the LTL carrier’s point of view.

Goods to be delivered start the day at a terminal location. Typically, these goods came in from other terminals overnight and were sorted to be placed on local delivery trucks for daytime delivery. After all, or in some cases most, of the deliveries have been made, pickups are made from shippers that are brought back to the terminal. If these goods are to be delivered to local customers, they are kept in the terminal; otherwise they are sent overnight to another terminal.

Typically, a truck is sent to some specific geography to do the daily work: deliveries first and then pickups. Some shippers have a schedule where pickups are made on a regular basis, some call in for pickups at least a day in advance and some call in the same day.

This type of route planning is fairly simple: build a route that puts the deliveries in a least cost delivery sequence, with the delivery at some point far away from the terminal and then stop for the known pickups on the way back to the terminal. Same day requests can be sent electronically to the driver along with a new route to be used for the trip back.

The problems with this approach is that the truck that is sent to a specific area may be underloaded or overloaded on any given day. When overloaded, the load can be split with the result being two less than efficiently loaded trucks going into the same area. Pickups can also provide the same challenge. Sometimes there are too many for a single truck and sometimes the truck can come back with very little.

An automated solution can be used to reduce costs.

When the work for next day is known (what deliveries are expected at the terminal for delivery and what pickups are expected), all of the work can be processed by an AI assignment algorithm to build routes for delivery and pickup.

The assignment algorithm can consider factors like whether or not a pickup can be made when a truck is nearly empty or that the limits of the trailers are not exceeded. If there is flexibility in the sizes of the trailers used, small trailers can be assigned to areas with less work, and larger trailers where there is more work.

If there is a need to assign specific equipment or drivers to specific areas, this too can be considered.

Can an existing transportation application be interfaced to an AI assignment process? It can be less complicated than imagined.

Strategic Movements has a suite of interfaces that can be used for this purpose. A robotic process interface can be used to move the data into Strategic Movements for assignment and other interfaces can be used to extract the assignments to sent to existing handheld solutions.

Looking for a better way to efficiently schedule deliveries and pickups? Visit us at Strategic Movements and see what we can do for you.

Like the blog? Sign up and be the first to hear what we have to say.