Definition – 6 MIN READ | 03-03-2020
What is route optimisation?
Route optimisation is a process which leads to ‘optimized’ routes. Route optimisation strives to build routes that are optimal in terms of the resources used and for the needs of a specific industry. Discover the challenges and a few definitions concerning route optimisation for transport and services.
Definition of “route” ;
The objectives of the route optimisation ;
Definition of “ordering” and “route building” ;
Definition of “route scheduling” ;
The algorithms of the route optimisation.
What is a route
A route is the path taken by a person or a vehicle from a point of departure and back to that same point. This itinerary can be punctuated by a set of visits or deliveries, pick-ups, maintenance operations, tests, collect of samples, etc.
Each of these actions may require specific skills of the field workers or a particular vehicle characteristic. Depending on the industry, it may be necessary to schedule some actions before others. For example, pick-ups followed by deliveries. Sometimes, each action must meet a constraint of time and duration.
What are route optimisation issues
Route optimisation focuses on ensuring the adequacy and optimisation between human and material resources in order to carry out tasks in different places (appointments, services, deliveries). Field workers leave from one place and return to the same place once the work has been completed. Route optimisation may be simple when there is only one worker in the field. It becomes more complex when there are several field workers and when several constraints are added to the tasks to be carried out and are placed on the available resources. It then requires the use of route optimization software.
This software has different capabilities depending on the requirements for use and the scope of the desired optimisation.
The use cases define the optimisation objectives and the desired levels of constraints. Most industrial use cases focus on two main objectives: the minimization of mileage and the maximization of the number of visits carried out. The usual constraints are the type and capacity of the vehicles, the availability and the skills or the accreditation of the technicians or drivers.
The scope of optimisation also affects the algorithms implemented. These solutions enable:
route scheduling only ;
route building ;
appointment scheduling and route building.
These latter solutions provide complete route optimisation for the entire business process.
What is “ordering”
The ordering is the action of placing tasks into a given order. These ordering algorithms are used when only a single field worker is taken into account. The objective of this order may be to minimize mileage. Nevertheless, it quickly becomes more complex when delivery hours are defined. Furthermore, when it is not a question of minimizing mileage, costs or the use of resources necessary to complete the order (installation/removal of a piece of equipment throughout a route).
What is “route building”
Route building is the fact of creating routes from among a cluster of points to visit, taking into account the constraints on the visits to be carried out, as well as the equipment and human resources. The solution obtained by route building is often not unique. Given the number of parameters and the acceptable time to obtain a solution, the optimum reached is often partial. Nevertheless, it is usually sufficient for professional use. The algorithms used in the context of route building are much more complex than ordering algorithms.
What is “route scheduling”
Route scheduling adds a dimension to route building. In the context of route building, optimisation focuses on a given cluster of points. On the other hand, route scheduling consists of taking a horizon into account and therefore considering several clusters of points and choosing one. Therefore, the question is whether it is preferable to perform a particular visit on a particular day or on the following day. The complexity of the matter is to know the impact of the addition of a visit on a specific day without building routes. This route building operation would be costly in terms of time and would not allow a quick decision to be taken. Specific route planning algorithms have therefore been developed and used to enable this quick response.
What is a “route plan”
A route plan is the result of the optimisation of routes for any given day. It consists of the itineraries of the different routes to be completed.
What is a “route optimisation algorithm”
“Algorithm” is a word that can frighten some professionals. Nevertheless, it reflects something fairly simple. An algorithm is a set of instructions given to a machine (computer) to perform a series of calculations. In the context of optimisation, the software program executes the same instructions several times, and at each iteration the result obtained will be closer to the desired optimum. An algorithm often responds to a given problem. Nevertheless, some algorithms may sometimes be easily adaptable. Finally, it should be noted that the development of an algorithm is very technical. The desired optimum must first be defined from a mathematical point of view. Then, it is necessary to establish acceleration strategies that will enable the optimum to be reached quickly.
For the route optimisation, the objectives and the methods of optimisation used may vary. They vary according to:
the constraints imposed by the industry ;
the customer promise ;
or the requirements of management.
The meaning of the term “optimized” varies according to the persons concerned and the industry. It is necessary to define the desired optimum.
Therefore, the search for a route optimisation tool starts with the definition of the desired optimum and the clarification of the constraints to be taken into account.
There is certainly a common basis specific to several typologies of professions or organizations in terms of objectives. In addition, route optimisation will each time take into account the constraints associated with human resources (availability, skills, accreditation) and vehicles (speed, capacity, etc.). Nevertheless, prior discussion is useful to check the adequacy of the optimisation algorithms present in the software “optimisation engine” with the required optimum.
For example, a few optima that are usually required:
Minimization of fleet mileage ;
Minimization of the resources used in order to maximize the number of visits performed ;
Prioritization of certain types of operation.
Aren’t all route optimisation algorithms known today?
The most straightforward algorithms are available in the literature today. In particular, logics of scheduling are established today.
In contrast, as indicated earlier, the process for the search for the optimum may be different for each problem. Furthermore, as the industrial optima being sought today are richer in terms of constraints, the research methods are therefore more complex.
Nevertheless, beyond the convergence of the algorithm towards a given optimum, the question of the speed of obtaining a “suitable” solution arises.
In the end, there are still many things left to discover in operational research, particularly when it comes to route optimisation.
For route optimisation: can we speak of artificial intelligence
Today, optimisation engines must use artificial intelligence in order to converge quickly towards the desired optimum.