Last-mile distribution has one simple goal: to move goods from a transportation hub to the final delivery destination. But if it were that simple, logistic experts and optimization software would not be needed. The most important factors when planning your distribution are the constraints and environment, the quality of service (QoS), and the cost.
Distribution optimizers, due to Artificial Intelligence, can generate more efficient routes. However, they must comply with the constraints to be feasible. In addition, there is a balance between the QoS and the costs that must be determined according to the goal of the company.
Constraints and Considerations
There are three types of constraints and considerations to take into account when planning your distribution:
Traffic and physical restrictions: moving within a city in peak hour can take 2x or 4x the time it would take without traffic. In addition, not all the vehicles can access all the zones of a city (e.g, a supermarket chain has vans, trucks and trailers, and some stores cannot be accessed with big vehicles). This information is crucial when planning your distribution.
Quality of service considerations: clients usually expect delivery within a time window, and some of them require specific vehicles or drivers. In addition, an important QoS metric is to deliver the goods as soon as possible. Not meeting these criteria could lead to a loss of reputation and customers.
Business constraints: the most obvious are the drivers’ working times, breaks, and making sure they work a minimum set of hours. There are also many heterogeneous constraints that differ between companies, and the knowledge of logistic experts is needed. Some examples are limitations while preparing orders in the warehouse, agreements with transportation companies to deliver in certain zones, or peculiarities while delivering to some clients.
Cost – Quality of Service Balance
The goal of a good route planner is not to reduce costs as much as possible. It’s objective is to reduce costs while producing feasible routes, and with a balance of cost and QoS that meets the company needs. Let’s analyze the possible results that a route optimizer can produce:
The very low cost zone represents the routes with the lowest cost but that, due to not complying with the constraints, are not possible to execute or produce a lot of late, low quality or missing deliveries. These should be avoided.
The medium cost zone represents routes that are feasible and comply with the constraints. In this zone, the standard scenario (A) is to keep the QoS while reducing the costs, due to the efficiency of the Artificial Intelligence. From that, there are two more possible scenarios: to increase QoS with a lower reduction of costs (B), and to decrease QoS to increase savings (C) , even though this one is not very common.
In the high cost zone, the increase of the costs produces only a subtle grow of the quality of service. The scenario D represents to keep the same cost, and use the optimizer to slightly improve the quality of service, which is usually not worth it.
Every company has different constraints and different goals in terms of what benefit the optimized routes should provide. That is why it is important not only that the software can take into account all such constraints, but also that people involved in the distribution process participate with their knowledge and experience to provide such constraints to the system and make it add the most possible value to the company.