«We are looking for startups experienced with data processing technologies and machine learning models capable of integrating very diverse sources», Juan Manuel Díez, Autoridad Portuaria de Valencia
11/11/2022. The Autoridad Portuaria de Valencia collaborates with Opentop in its I Acceleration Program proposing the challenge «Prediction of daily port traffic«, which aims to predict daily port traffic based on the application of digital technologies, integrating data from multiple sources.
With this purpose in mind, we have chatted with Juan Manuel Díez Orejas, Head of Strategic Planning and Innovation of the Autoridad Portuaria de Valencia, who has given us more details about this challenge.
Currently, interested startups can apply with their solutions to this and any of the four challenges that the call has open before December 4, 2022.
What opportunities could this challenge have in the market and could it be applied to other non-port logistics sectors?
The combination of different data sources to predict traffic behaviors can be useful to anticipate scenarios in those infrastructures of modal interchange (such as airports) or where flows from different origins and destinations have to be combined and where there may be external factors influencing such flows.
What role should innovation play in meeting this challenge?
Identifying the databases that could influence the different flows, combined with the constraints that the system itself incorporates and being able to model the algorithms that would allow them to be correlated is innovative, at least in the port logistics field.
What kind of startups are looking for this challenge?
Startups familiar with data processing technologies and with the creation of machine learning models capable of integrating very diverse sources, both with structured and unstructured information (statistical databases and «sentiment» from social networks, for example).
Have there been any previous attempts to solve this challenge, and how?
There have been some ideas but, as far as I know, no concrete solution has been developed, not even at the prototype stage. Work is being done on structuring port data, but the actual scope would include DBs outside the conventional business scope of ports.
«Daily port traffic prediction challenge» by Autoridad Portuaria de Valencia
Do you want to know more about the challenge of Autoridad Portuaria de Valencia? Let’s review its history.
Ports are key drivers for the economy and trade and have an enormous impact on our personal well-being and on the competitiveness of the business fabric in their areas of influence. Every port must carefully plan its capacity to be able to manage the traffic of goods that are loaded, unloaded, and stored in the port. Many planning decisions, especially regarding infrastructures, usually are made based on the expected traffic and must be made long in advance so that those infrastructures are in place when they are needed. These decisions are typically taken based on multiannual strategic plans, in which preparation, traffic forecasts are made.
This pace was fairly enough in the past, but the demand for logistics services is experiencing changes that require faster decisions and therefore, updating these forecasts more frequently. In recent times, there have been several world events that have had a significant impact on port traffic, so it would be convenient to add their effects to the forecasts. In addition, given the increase in traffic and the saturation of infrastructures, it is increasingly important to have good forecasts also in the short term, that allow to plan the required equipment and workforce in any event to cope with the demand.
This challenge aims to obtain knowledge and forecasts (from both internal and external information) in order to be able to make more precise decisions and to be able to anticipate problems that may arise due to the evolution of traffic. Obtaining forecasts should be an agile process and should consider also new variables that are not currently considered but which have an impact on port traffic such as the number of TEUs, the scale of vessels, the number of vessels and land operations (truck and rail), etc. All of this, in order to be able to have new frequent forecasts (daily and differentiating between mornings and afternoons) giving the possibility of reacting and recalculating them when events occur that modify traffic conditions.
What are we looking for?
We are looking for startups that can combine data from different sources, like market data, social networks sentiment, and innovative new data sources, and apply technologies like AI and Big Data to provide accurate and fast predictions of port traffic.
This is a great opportunity for the sector, and we hope to generate interest from the most talented startups!
We invite you to take a last look at our acceleration call and register now, we are waiting for you!