Awake.AI from Turku won the tender to provide vessel schedule estimates for Finnish ports through Fintraffic Vessel Traffic Services. The new port call schedule service will improve the competitiveness of Finnish maritime logistics, boost the efficiency of port operators’ routine activities, support the development of automation, help to anticipate exceptional circumstances, and reduce environmental emissions.
Fintraffic´s Vessel Traffic Services has signed a service agreement with Awake.AI for the new Port Call Time Stamp and Estimation Service, that is, a time data service for port operators and authorities. An EU-wide tender for the public procurement attracted a great deal of international interest. The technical implementation of the project will be carried out in Finland. The new port call schedule service will improve the competitiveness of Finnish maritime logistics, boost the efficiency of port operators’ routine activities, support the development of automation, help to anticipate exceptional circumstances, and reduce environmental emissions.
“The importance of time data for shipping has been taken into account in both national transport system plans and the government’s decision-in-principle on the digitalization of logistics. The port call schedule service will now enable us to implement this on a practical level,” says Olli Soininen, Programme Manager at Fintraffic´s Vessel Traffic Services.
Awake.AI was able to offer a ready-made service that successfully met both Fintraffic’s needs and the requirements set in the invitation to tender. The service can therefore be made quickly available to port operators, as development work does not need to start from scratch.
“In practice, the new port call schedule service will provide the necessary APIs from which port operators can obtain time data for use in their own systems,” says Kimmo Kummala, Project Manager and Vice President of Engineering at Awake.AI.
Portcall schedule working group indentified practical requirements
The new service owes its existence to the port call schedule working group, which was launched in spring 2019. The group was led by Traficom and consisted of about 30 operators in the marine logistics sector: companies, public-sector entities, organizations and authorities that use and produce time data.
“There have long been attempts to tackle the challenges posed by data management in maritime logistics. Before the working group was established, a preliminary study was carried out to collate existing research and reports and to identify any overarching needs. Time data clearly took centre stage,” says Katariina Kalatie, Chief Advisor, Ship Technology and Marine Environment at Traficom.
The preliminary study revealed a largely general consensus in terms of needs and the target state: better predictability and information sharing is required to boost operational efficiency. The challenge was how and what data could be shared openly, as some information contains trade secrets and many operators are in competition with each other. Time data was found to be something that would benefit all port operators equally.
“The members of the port call schedule working group make practical use of this information in their daily work. This gave us a good idea of what was really important and relevant at a grassroots level. The goal was to find practical solutions rather than high-flying strategy slogans,” says Kalatie.
Artificial Intelligence learns from the past and adjusts estimates accordingly
The arrival and departure of a ship from port determines the schedules of countless people. Although the majority of vessels have had to provide an estimated time of arrival via VHF frequencies in addition to other necessary information, the data obtained from the vessels has been poor and fragmented.
The new service will be based on data analysis through machine learning. The forecasts will be influenced by many factors that will be analyzed using global AIS messages. In addition to automatic classifications, such as where a ship is coming from and where it is going, a ship’s estimated time of arrival will be influenced by many variable factors, such as its speed and route, the weather and the ice situation.
“The AIS messages sent by vessels hold a lot of potential, but for some reason, their benefits have so far been ignored,” says Jussi Poikonen, Awake.AI’s technical project manager.
By utilizing machine learning, the service will remember what has happened in the past and how various factors have affected a ship’s arrival time. Artificial intelligence will therefore learn the normal deviations for a particular locality and adjust its calculations accordingly. This will clearly reduce data fragmentation and generate more accurate schedules.
“Although price was naturally a factor in the tendering process, it was great to see that due attention was also paid to data quality. Fintraffic has clearly taken the value of data into account,” says Sami Kaksonen, Vice President of Sales and Marketing at Awake.AI.
Port call schedule service to be introduced in record time
“The new port call schedule service will be introduced this autumn. Now that the agreement has been signed, Awake.AI will begin an approximately month-long delivery process during which the service will be technically customized to meet Fintraffic’s needs. This will be followed by a test phase, after which Fintraffic can mobilize the service for use at ports. The service will be available through the Fintraffic Digitraffic API or the Fintraffic Port Activity app,” says Juho Pitkänen, Development Manager at Fintraffic´s Vessel Traffic Services.
The service will first be introduced at Finnish ports, but can in practice be used anywhere. This offers global opportunities in a variety of sectors. Finland is a world leader in digital development, and activities such as the port call schedule service will ensure that Finland remains at the forefront of global digitalization.
“Fintraffic has handled this project superbly, and shipping will now take a great leap forward. We hope that the service will be actively further developed to meet the needs of various operators,” says Kaksonen.
“This is a great example of networked collaboration in which each party has supported the project at different stages. I hope this can be developed on a long-term basis, and that more operators will get involved. Better predictability is of concrete benefit to many people’s everyday work, and I hope that up-to-date and reliable schedule forecasts will be available to everyone who needs them within the next couple of years. Perhaps we can also find other similar areas for development that can be improved with the aid of data”, says Kalatie.