Courier vs machine, and the winner is...

In addition to a TMS system, couriers are increasingly being assisted by route optimization software. So... does that work? Who is faster? The carrier, or the software?

‍Unequal competition

‍First of all, without a courier, the order will not arrive (yet). Despite all drones and self-driving cars. In addition, comparing the carrier with and without route optimization software always consists of two unequal matches. Unfair to both parties. The first race is to optimize the ideal route at the start of the day without knowing any exceptions or ad-hoc assignments. The other is that of the resulting route. Since a lot always changes throughout the day, the latter is the real comparison. You can also improve that route by means of software. Both the courier and software each have their own advantages. An experienced courier knows his area like no other; shortcuts, parking options, appointments with customers, expected crowds, etc. These are all examples of information that the software does not have available (at the beginning of the day). This also includes the business park from which bookings come in every afternoon. Prior knowledge about this means that the courier plans its route early. This terrain comes at the end of the day. Choosing the theoretically fastest routes along all stops in a day can make route optimization software much more accurate and faster than a person can do. Digital maps contain all the knowledge of the road network and any (historical) traffic. Keeping this information up to date for the entire country is a piece of cake for the computer but impossible for an individual.

‍Best of both worlds

‍Combining this knowledge increases the efficiency in the process. People are intuitive, so they are often better aware of relevant circumstances. And can often predict better based on experience. Especially if it has not yet been defined in systems. But software is better at optimizing routes and has more overview. Nowadays, humans are losing it to the computer. When Google says to me, “You have to get rid of this to avoid the traffic jam”, then I go off. The chances of me knowing better because I have all the insight are almost nil.

‍Combining knowledge

‍Ultimately, you want both the courier and the system to have the same knowledge at the start of the day, which is why historical data is important. When the system learns from the couriers, it gives advice to new couriers later. When a system knows you should stop at the back of a company, it includes this information in future routes. Unplanned pickups are generally disastrous for your route planning. From historical data, you can now deduce in which area many pickups occur at what time. The system can take this into account in route planning by placing these areas later in the day. You combine the knowledge and properties of both sources of information in order to arrive at an efficient and optimal route for the courier.

‍Preserve freedom

‍The courier must still have the freedom to deviate from the routes. During the ride, he controls the wheel and the system is supportive. In the event of an alternative choice, the system must recalculate the route and provide new advice to the carrier. Experienced couriers will notice the influence of such a system to a lesser extent. Inexperienced couriers, in particular, will drive more efficiently. There is no need for familiarization and the efficiency in the case of an “inexperienced courier” is immediately high. In addition, the flexible deployment of couriers in unknown areas is possible. This allows you to easily scale up during busy periods to meet customer demand. This is now finally possible without sacrificing efficiency.

‍The reality

‍CGI data analysts have researched for DHL to improve routes by using historical data. This study showed the relevance of using route optimization software. This shows that in 84% of the cases, the software calculated the fastest travel time and shortest distance. In practice, this also resulted in 10% more deliveries per vehicle, and in an enormous reduction in induction time. In 5% of cases, the courier is faster due to a different route planning. In the other cases, it is a tie. The study also showed that there are sometimes long stopping times. In some cases up to 55 minutes. The use of historical data may well help to plan stops so that waiting time is minimal. Or at least, to provide clarity about this. Overall, there is still an enormous amount of profit to be made. In short, don't see route optimization as a replacement for your courier experience, but as an addition that builds on your couriers' experience and knowledge. So in the battle of man and machine, there is another winner... and that will be the customer.