Last mile delivery is known to be the most complex and costly part of the supply chain. In recent years, the growth of e-commerce and the increase in the number of retail outlets in urban centers have further contributed to this trend, multiplying the volumes of packages to be delivered. Moreover, the increase in delivery services makes consumers (and professionals) more demanding than ever and less willing to pay for delivery.

What is true for parcel delivery extends to other activities. Long delivery times or missed deliveries are no longer acceptable to customers, whether it is to receive a parcel, a delivery, a technician or a health professional.

These underlying trends make the last mile increasingly complex to manage for organizations that now need to be able to plan it very carefully.

Route optimization is the best solution! But…

By assigning the right resource to the right place at the right time, it allows you to better meet your commitments to your customers (especially in terms of time slots), to reduce your costs (fewer vehicles used, fewer kilometers traveled) and to decrease your environmental impact.

However, in practice, few companies are actually using such a solution. It is estimated that only 20% of the players who have this need rely on a route optimization software. Among those equipped, the majority does not use the solution at all or only in part, making a large number of manual adjustments and rescheduling.

Where does this paradox come from and what are the obstacles to using route optimization softwares?

Data quality is key

As with many technological tools for supply chain optimization, route optimization is essentially based on data. The addresses of the locations to be visited, the size and weight of the parcels, the time constraints of the delivery sites: the data defines the problem and the quality of the data determines the relevance of the solution that will be provided.

In some organizations, some information is known to the operational staff but does not exist in any database: this is sometimes the case for companies requiring technical intervention scheduling, incompatibilities between service providers and customers, or implicit agreements rooted in common habits.

As a result, many players who could optimize their routes do not do so because they do not have the capacity to provide the solution with sufficiently complete and reliable input data.

Logistics never stops and information is constantly added to the system. The idea of having to run a new optimization after each change (or to manage them manually) can put off potential users.

Logistics never stops and information is constantly added to the system. The idea of having to run a new optimization after each change (or to manage them manually) can put off potential users.

The latest smart geocoding solutions use technologies capable of working on unstructured data thanks to Deep Learning, significantly improving the quality of the results provided to route optimization solutions.

Taking into account specific constraints, details that change everything

Many activities have to meet very specific constraints. Here are 4 examples:

• furniture delivery requires 2 delivery men when the shipment exceeds a certain weight,
• some interventions require technicians to go to a PUDO (Pick-Up Drop Off) center to pick up equipment,
• the time spent by a technician on a customer’s premises can vary by a factor of four depending on the technical complexity of the intervention,
• in bulk shipping, some collection locations can only accommodate one vehicle at a time

If these constraints are not taken into account when planning the routes, the solutions provided will be neither adapted nor realistic, thereby requiring the user to make a large number of manual adjustments to the resulting schedule, further complicating the use of the software.

The ability to handle very specific constraints must therefore be a fundamental criterion when choosing a route optimization software.

Real time increases the gap between theory and practice

Traffic jams, absent customers, wrong address, parking issues: the last mile is full of obstacles. While most route optimization software solutions are static (route calculation and cut-off), the last mile is very dynamic and requires utmost proactivity!

How can you ensure that a traffic jam at the beginning of the day does not impact a tight time slot planned later on? How to rebalance the technicians’ rounds in real time if the schedule of one of them has become lighter?

The significant differences that companies observe between the performance of planned rounds and those actually carried out lead to disappointment towards route optimization tools. In the context of urban logistics it is becoming less and less relevant to calculate static routes.