Route Optimization Software Algorithms Defined
Route optimization software algorithms are the mathematical formulas that solve vehicle routing problems. They primarily are designed to increase efficiency, whether that be to reduce costs or reduce total time. An efficient set of delivery routes would have the least aggregated time, which should correspond to the least total cost. However, some vehicle routing problems are not primarily “driven” by reducing costs. In the case of disaster response, the objective is to reduce elapsed time to have all of the service objectives completed. For example, our routing expert, Kevin Callen, built the CDC’s Strategic National Stockpile disaster response routing system by overriding a typical cost-minimizing routing algorithm for point-to-point routing with parameters that required all of the routes to be completed faster and close to the same time. The same issue must be addressed in snow removal routing. All of the streets must be finished with being plowed as fast as possible and at roughly the same time.
Route Design – Vehicle Routing Problems Are Not All The Same
Vehicle routing problems are not all the same. Collecting garbage requires a different type of route design than delivering furniture or shuttling around elderly or handicapped citizens. Route design is primarily composed of three attributes: Schedule, Boundary and Sequence.
The Schedule is the day or time window of service (visit, delivery or collection). Boundary is the geographic extent of the route. A route boundary may be condensed and all stops are required to be contiguous. For example, a waste collection route or meter reading route will all be in one concise area to reduce travel time between stops or customers (and make it easier for the crews to remember).
A route boundary may be spread out over an entire region or state, such as for sales forces or infrequent onsite service providers.
Sequence refers to the order that the stops are visited in the route. In cases of high density routing, like waste collection routing, the sequence would seem very important. In reality, the drivers have a multitude of preferences and safety issues that the route optimization software does not know. Thus, the sequence for waste routing is not critical, but the schedule and boundary is critical (click here to learn more about waste collection routing). However, for furniture deliveries, the sequence (and schedule) are critical and the boundary is not important.
Types of Vehicle Routing Algorithms
The academics have a multitude of names the various types of vehicle routing problem algorithms. For us laymen and operational people that implement the route optimization software applications, it is easiest to categorize them into three categories:
High Density Routing
High Density Routing refers to waste collection, postal deliveries, snow plow routing, newspaper deliveries, and meter reading. High Density Routing is where you have a large number of stops (more than 200) serviced on the same schedule and in a contiguous boundary.
Point-to-Point Routing refers to collections, deliveries and sales forces with generally less than 200 stops per route per day. With Point-to-Point Routing, the boundary is not as important as the schedule and sequence.
Paired Routing refers to para-transit, armored car routing and similar. Paired Routing is similar to Point-to-Point Routing, but the sequence requires the stop to be “paired” with a facility trip. A elderly person is picked up by the shuttle and taken to the shopping mall and must be returned to their home. In a normal Point-to-Point Routing problem, the start or end point is the same for many or all of the stops.