How to solve the great delivery driver shortage

German Autolabs
5 min readFeb 17, 2021


The parcel industry is exploding, but logistics and delivery companies can’t keep up with demand. That’s why German Autolabs is developing AI-powered language tools to help recruit and train new drivers. CEO Holger G. Weiss explains all.

The CEP industry (courier, express, parcel service providers) faces a dilemma that many industries confront when they grow disproportionately. Developments in recent years, and especially since the Corona crisis, have driven up the volume of parcels shipped month-on-month. The figures are staggering: at the end of 2020, Deutsche Post alone estimated that it would have transported 1.8 billion parcels in the whole year — in Germany alone. The total volume for Deutsche Post was close to 4 billion parcels. This immense growth can only be managed through massive investments. New distribution centres are being built, automation is advancing, and robots and drones are helping.

Of course, the parcels also have to be distributed, and it becomes clear to anyone looking at a typical German city centre that the increase in parcels has also led to an increase in vehicles. In some streets you can see multiple yellow, red, blue or white vans. One can only imagine the consequences for traffic and cities. But this is also where the industry comes to an abrupt limit in its growth. Logistics sorting plants can be built, robots can be set up and van manufacturers can ramp up production. But qualified drivers who do the job of delivering millions of parcels every day within a certain time window cannot simply be cloned. Automating and optimizing the last-mile delivery process will remain a field of experimentation for decades to come.

Since 2007, the number of workers in the CEP sector has increased by more than 50%, and a significant percentage of these workers are drivers. In an article on their website, the provider NOX Nachtexpress even admits that in the long run, a shortage of drivers could jeopardise security of supply. In addition to the challenges of finding new drivers, the industry also struggles with other challenges. The working environment has suffered for years from a lack of reputation, as well as stress and poor pay. These are not ideal conditions in which to find candidates and keep them. The industry talks about 25% — 30% churn rotation of drivers, which in turn leads to vacancies and significant costs to train the next new driver.

The problem is well known, but it seems that the philosopher’s stone has not yet been found in terms of fast and efficient driver recruiting. Large companies in the CEP sector depend on subcontractors to provide vehicles and drivers of a certain quantity — while guaranteeing baseline quality. This is a major problem. It is not enough to find someone with a driver’s licence, they also need to be trained for a route, pay attention to all the details, and understand that the training can take weeks. In the industry, companies prefer to pay high prices to subcontractors rather than face this problem on a structural level.

Yet there are solutions that are breaking new ground when it comes to substantially improving this system. At German Autolabs we have developed a technical solution that allows drivers to be supported ‘on the job’ from day one. The service is a context-sensitive voice assistant that acts like a co-driver and steers the driver around all the usual pitfalls automatically. The app uses information from route planning, but can also integrate other sources. How many parcels need to be taken out of the car at the next stop? Oversized? Are there special instructions such as identity verification or a previously agreed drop-off location? The assistant announces the necessary information when it is needed. In doing so, it uses the GPS position and can even anticipate if there is a deviation in the sequence.

German Autolabs’ Last Mile Assistant graphical user interface (GUI)

The system is multilingual and thus addresses another aspect: many of the drivers come from foreign countries, often in Eastern Europe, and the more they understand, the better and more productive they are. As a rule, the system is installed on a tablet directly on the windscreen and runs parallel to the scanner. The latter always remains ready for use. When the driver feels confident enough and has internalised the route, the system can simply be removed and used in another vehicle. The voice assistant is always synchronised with the route via a link to the scanner API (application programming interface).

In a recent case study we undertook in England, in which one of the big players in the industry tested the system under real conditions for several weeks, it was shown that new drivers sometimes reached 100% of the route load in less than half the usual time. The system thus addresses not only the urgently-needed digitalisation of drivers’ workplaces, but also very specifically the costs associated with training new drivers, as well as the quality of service. This is a big step towards a sustainable improvement of the drivers’ performance and the biggest problem that the entire CEP sector is currently facing.

For full access to the results of our comprehensive field test, and to find out more about how we build voice assistance for logistics and delivery fleets, visit — and as always, thank you for reading.



German Autolabs

We build logistics voice assistance for mobile workers.