The human brain has about 100 billion neurons that are interconnected. Information is transmitted between those neurons in the form of electrical impulses. This enables us to learn, draw conclusions and think abstractly. As for so-called artificial intelligence, “neurons” that are simulated artificially and trained by means of algorithms are used. However, the goal is not to reproduce human intelligence. Instead, machine learning enables systems to learn pattern recognition based on a large amount of data.
For instance, with machine learning, a system can automatically recognise a set of rules based on training data. Thanks to machine learning, companies no longer need to create models manually, which means that they do not need to spend time on defining rules, checks and interpretations anymore. The quality of the training data is decisive for the successful application of machine learning.
Two tasks are particularly challenging when it comes to the development of a machine learning model. One of them is the so-called feature selection, i.e. the selection of a subset of relevant properties of a data record from the numerous properties of past transport orders. For example, the selection of the destination, weight or transport type. The second challenging task is the so-called “overfitting/underfitting”. From a mathematical point of view, the model needs to be complex enough to learn human behaviour. However, it should not just memorise that behaviour. The desired solution is referred to as a “generalisation model” by machine learning engineers.
The logistics software CarLo, developed by Soloplan GmbH, has already been operating with artificial intelligence for over a year. Machine learning is based on transport planning data, such as the shipment mode, date, start and end point, load items or weight of the load. This information is fed into the system and processed with the help of an algorithm.
One of the newest extensions in the AI functions of the transport management system CarLo is a prediction for delays or the reliability of drivers and subcontractors.
Another new function in CarLo’s AI module is the provision of tour suggestions. Based on experience and planning behaviour, the system learns to plan tours effectively. The AI takes the Quick Steps configuration already existing in CarLo into account while putting together tours. This unique combination of probabilistic and deterministic functions allows our customers to define restrictions such as the maximum weight or loading metres. CarLo’s AI will only make suggestions that meet the defined restrictions.
The Soloplan AI team is currently looking into a number of other ideas. A topic relevant to the sector is the ordering probability. How likely is it that an offered transport will actually be ordered? This question and other interesting topics are currently being evaluated; it is expected that these functions will be available for all Soloplan customers in the future.
The advantages of using machine learning for transport planning are obvious: It saves the dispatcher a lot of time, helps avoid mistakes and increases efficiency. Another advantage is that knowledge is no longer lost in the case of personnel changes. Since the transport management system learns the required behaviour based on training data, a new dispatcher, for example, can plan tours in the same way as a long-term employee. All data remains with the customer at all times. Furthermore, the pipeline adapts to changing business requirements as the model is further trained with new transport orders.
GO DIGITAL! – The digitalisation initiative of Soloplan GmbH
In the scope of the digitalisation initiative 2020, Soloplan shows you innovative IT solutions and functions from CarLo that you can use to digitise your analogous processes: OCR scanning, voice control, workflows and much more!
You would like to learn more about the logistics software CarLo?
Then feel free to arrange an individual presentation appointment today via e-mail to email@example.com, by phone at +49 831 57407 300, or using our LIVE chat at www.soloplan.com.