Artificial intelligence instead of trucker romanticism
In the logistics industry, supply chans are still often created by hand. The casual charm of the trucker world can be particularly tenacious in commercial transport. A start-up from Hamburg is now shaking up the scene by employing artificial intelligence (AI) to ideally distribute and control goods. Carrypicker founder Andreas Karanas explains in an interview how carriers, customers and even the environment can benefit from this software, which was also awarded the Weconomy Founder´s Prize.
Mr. Karanas, would you consider Carrypicker more of a software smithy or a forwarding company?
The core of our product is the software programmed by us, which utilizes AI to optimize the flow of goods. This is one of the reasons why the majority of our employees are mathematicians and programmers. But to the outside world, we act as a forwarding agent. We are not an intermediary for individual transport journeys, but instead we cover the complete risk of transport and organize the necessary transport capacities.
And without your own trucks, right?
That is correct, we purchase transport capacity from our partners, which now include around 2000 transport companies. The difference to traditional forwarders is that we complete part of the planning process in an entirely digital manner. Dispatchers often work using their vast experience and with planning tools or Excel spreadsheets – our AI runs in the background. It sorts, plans and organizes. It is like Tetris: Ideally, we group cargo from several separate consignments and then determine a price for each individual freight. The trucking companies commissioned by us then carry out the actual transport.
Trucker romance sounds different
Our digital approach is new in this industry and was initially met with much skepticism. Now, however, many have recognized that capacity can only be increased with the help of artificial intelligence.
Forwarders cannot provide the expertise necessary for this, which is where we come in and realize the necessary digitalization. The trucking company no longer has to worry about how their trucks are utilized, because we take over the entire vehicle and consolidate the goods ourselves. Smaller companies especially appreciate this, as it is a service that they would often not be able to implement on their own. Now they benefit from faster orders and improved capacity, which is ultimately reflected in higher earnings.
What can your AI do better than a traditional dispatcher with decades of professional experience?
A traditional dispatcher works just like us: Orders come in and they are bundled. However, our AI can take many more factors into account simultaneously. We have identified around 50 aspects that are relevant to our calculations. These include obvious factors such as size, weight or, of course, distance. However, there are also other variables that do not have a great effect on their own, but together can have a major impact.
What sort of variables?
Weather conditions, possible public holidays or special circumstances at the destination, which in turn could impact the loading options for the return journey. Once all properties that influence the transport of goods are included in the calculation, it becomes extraordinarily complex. Thanks to years of experience and intuition, good dispatchers achieve capacity utilization of just over 70 percent. With our AI, we can reach 80 percent. Another advantage of the software is that it is faster and can analyze a greater amount of data at once. One person can only plan a small number of cargos per day; however, the software gets better and better the more data is available.
What are the advantages that artificial intelligence offers trucking companies?
The price war in this market is fierce and does not offer trucking companies much room for maneuvering. Large cost blocks, such as the price of diesel, are the same for all competitors. In order to significantly increase margins, the only thing that really remains is improved capacity utilization. Using traditional methods, however, the capacity cannot be increased any further. This can only be achieved with disruptive technologies – such as AI. A computer does not care if it has to calculate four million possibilities or 40 million possibilities as long as the calculation model has been programmed properly. This is only limited by computing capacity, which can be easily hired nowadays.
Who is your typical customer, and can the software also be used in other areas?
Our software is tailor-made for this one area of application. Of course, the problem of the last mile can also be solved algorithmically, but that would require entirely different software. However, when it comes to bundling larger flows of goods, there are various possibilities. This enables us to optimize the supply chain for major customers, for example if parts from several locations are to be combined. We know how to combine routes, times, demand and prices to form an ideal relationship – especially when other flows of goods are included in the calculation. Our customers therefore also include very large corporations, with an emphasis on larger medium-sized companies that have a consistently high transport volume and want to manage this as efficiently as possible.
What downstream positive effects can you achieve by better utilizing commercial trucks?
First of all, less transport capacity is needed for the same amount of goods. Fewer vehicles also mean, of course, less traffic on the roads and lower exhaust emissions. The environmental impact is so significant that large companies approach us wanting to improve their sustainability balance sheet.
What scale are we talking about here?
Thanks to an initiative from the non-profit organization GLEC (Global Logistics Emissions Council), we can precisely calculate CO2 savings and have received a seal for this. According to current statistics from the German Federal Environment Agency, truck pollutant emissions per ton kilometer average 103 grams of CO2. The GLEC took the data from the German Federal Environment Agency and simulated how major the savings effect would be with our software technology. The result: With our algorithm, we can achieve 95 grams of CO2. Extrapolated to the total commercial truck traffic of 600 billion ton kilometers in Germany, that amounts to 4.1 megatons of CO2. For comparison, this would be about one tenth of all savings measures implemented in Germany last year.
The Federal Ministry of Transport has therefore funded Carrypicker with 2.5 million euros.
Yes, because the Ministry of Transport does not want to fund just
one individual company, but is instead interested in supporting
something that could drive Germany forward as a whole.
It’s like Tetris: Ideally, we group cargo from several separate consignments and then determine a price for each individual freight.andreas karanas, founder and managing director, carrypicker gmbh
Fewer vehicles also mean, of course, less traffic on the roads and lower exhaust emissions.andreas karanas, founder and managing director, carrypicker gmbh
WECONOMY brings together innovative, technology-oriented start-ups and established companies – to turn great ideas into successful companies. The Gründerinitiative (Founder Initiative) is a strong network within the German economy, which offers access to valuable contacts and many opportunities for open and intensive exchange with experienced specialists and executives. The competition is organized annually by Wissensfabrik – Unternehmen für Deutschland e.V. in cooperation with UnternehmerTUM, Europe’s leading center for innovation and start-ups, and Handelsblatt.