Urban logistics is facing increasing challenges: B2C customer requirements are becoming the norm and are greatly increasing the complexity of logistics, which have to satisfy more demanding requirements in a context where all players are urged to be more environmentally friendly, where cities are increasingly pedestrianized and where the shortage of delivery drivers is putting the entire sector under pressure.

Given these challenges, route optimization technologies are the answer to the problem. However, many players who have tried to use such tools on a daily basis have experienced major disappointment.

One of the reasons they give is that the reality is too complex for these tools to be of any real help. Isn’t this a paradox for tools whose purpose is precisely to manage complexity?

The root problem: reality is unpredictable

Because urban logistics is by nature a changing environment, it must deal with constant changes that make it difficult to estimate routes in advance. Indeed, no matter how detailed the data may be, it can only account for the great diversity of situations that a driver may encounter on the road and that the same route or the same delivery can take twice as much time from one day to the other.

These differences are caused by two types of factors: internal and external.

Internal factors

The internal reasons are those caused by the logistics operator’s own activity. No process is ever perfect, and many hazards can disrupt a round:

• a late or absent driver
• a new driver who does not know the area well
• a incorrectly weighed or measured parcel that does not fit in the vehicle
• a parcel lost in the warehouse
• a breakdown or breakage that makes certain deliveries impossible
• etc.

If it is possible to limit the impact of these hazards, it is easy to understand that it is impossible to avoid them completely. Moreover, these events are out of proportion in number and impact compared to the external factors.

External factors

These types of hazards depend on external factors, to which the operator can only adapt. Some of these include the following:

• road traffic, road accidents, closed off areas due to demonstrations / road works
• a customer who is absent or who takes a long time to meet the delivery person
• a change in the delivery time slot during the delivery round
• poorly written, imprecise or a wrong address
• difficulty to find the entrance
• no parking space available
• etc.

However, in practice, operational teams and especially drivers, adapt to these difficulties with incredible resourcefulness on the road. Thanks to their detailed knowledge of the area to be delivered, drivers avoid delivering a street with a school between 4pm and 5pm because they know that they will not be able to park, they know that some hotel deliveries are made at the back via a one-way street, they group the deliveries for a same neighborhood and do “micro-rounds” on foot in the densest Parisian neighborhoods, etc. There are endless examples of the resourcefulness on how drivers adapt to change.

Just like unexpected events that occur during a round, a driver’s inputs represent endless changes to the route initially calculated.

The solution: route optimization that continues even after drivers have left the warehouse

Today the majority of route optimization softwares work in a static way, i.e. they calculate the routes, which are sent to the drivers on their PDA (or sometimes even printed out on paper before drivers start their round).

The most advanced solutions track the round through the PDA and recalculate in real time the ETAs (estimated time of delivery) according to the progress of the round. Some solutions also allow the planner to send new assignments on the go, which are then intelligently added to the ongoing round. However, this is not enough: 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 can we rebalance the pick-ups in real time for the whole fleet if one of the drivers is ahead of schedule?

The reality of urban logistics is so changing that not taking into account (or leaving the driver to manage it alone) all the uncertainties that will occur during the day is no longer acceptable.