Sustainability and greenification are no longer nice to have and are becoming strategic topics in every boardroom. This has been exemplified for the maritime and ports industry from a range of past events where the industry is now moving in the right direction and starting to leverage geospatial data and artificial intelligence to realize the goal of a green port.  

On November 17, 2021, cargo ships queuing up at Los Angeles and Long Beach ports were forced to wait 150 miles outside of the port so emissions from idling ships close to shore would be diluted before reaching the coastline. Usually, they would idle at around 25 miles outside the port, as shown in Figure 1. Knowing that cargo ships sometimes wait multiple days before entering the port, there is room for improvement.  The xyzt.ai platform enables you to instantly identify these points of improvement by visualizing your entire port operations and providing interactive tools to analyze the situation in real-time or based on historical data.

Environmental issues and new regulations

Within major ports, degassing of tankers, where remaining vapors are removed from the tankers hold between the unloading and loading of two different liquid cargos, is still happening when in transit resulting in high levels of pollution and annoyance for nearby residents.  

New regulations are being implemented to force a greenification of vessel fleets, such as the Carbon Intensity Indicator (CII), a rating for the vessels annual carbon intensity. From January 2023, all vessels that trade internationally and above 5,000 gross tonnages will be subject to new CO2 regulations. To comply, ship owners and operators will have to monitor the performance of their fleet closely and pro-actively to ensure regulations are met. 

Thanks to digitalization and innovation in the maritime and ports industry, the future is looking bright. Vessels are continuously tracked, and engines and performance are increasingly monitored using sensors, IoT devices, and cloud connectivity. Ports use radar systems, AIS receivers, cameras, environmental sensors, and digital twin technologies to obtain situational awareness to detect and react in real-time. 

Geospatial data and AI for a greener future

Geospatial data plays a pivotal role, and AI is increasingly becoming the engine fueled by the data to accelerate ports and the maritime industry to a greener future. 

Take the example of vessels queuing at ports. The historical data and the current (global) environmental picture can predict wait times and congestion. They can be accounted for by letting vessels cruise at a slower pace, adapting their ETAs, hence burning less fuel on a voyage and when idling. This is a global optimization problem, but the data is there through services such as Spire Maritime’s satellite AIS data set. 

Or the example of illegal degassing. The pattern of a tanker moving between two locations and trying to get rid of its hold is easily detected using the geospatial location records obtained from the radar or AIS system in the port. This was showcased by Port of Antwerp and xyzt.ai using six months of historical vessel data and unsupervised machine learning. Unsupervised means the algorithm was trained from the data only, without any user intervention necessary to annotate prior data. 

Port of Antwerp and xyzt.ai

Algorithms fused radar, AIS, and geospatial data about the location of terminals, locks, etc. xyzt.ai used data-driven algorithms based on k-nearest neighbors, dynamic time warping, and deep learning to identify anomalous vessel movements in Europe’s second major seaport. The algorithms could extract illegal degassing events from hundreds of millions of journeys of historical data. An example is shown in Figure 2, where a tanker performs a “sight-seeing” tour between two close-by stops. 

The collaboration between Port of Antwerp and xyzt.ai to detect anomalous vessel movements and accelerate the transition to a smart and safe port is illustrated in a joint webinar. Sign up to re-watch the webinar here.

European Copernicus data and atmospheric data

The impact of these improved algorithms can be monitored and analyzed using geospatial data such as atmospheric measurement data from, for example, the European Copernicus program. Daily measurements can be visualized and analyzed and used as input data for predictive AI models. Figure 3 shows how nitrogen dioxide (NO2) pollution data can be analyzed in the xyzt.ai platform using complex queries answering questions such as “Which regions saw the most days with high NO2 levels?”. From the image, it is easily seen that ports such as Port of Rotterdam and Port of Antwerp, the two largest European seaports, significantly impact pollution levels. 

Conclusion

The maritime industry and ports worldwide face a significant challenge to reach the increasingly stricter environmental regulations. At the same time, competition for maritime ports is more challenging than ever, and ports need to invest and quickly adapt, not to be qualified poorly or become constrained by the regulations for future expansion.