Do we really know what a “smart” port is?


Whilst the title of the article is mainly provocative, it is not very far from reality. Eager to keep up with the latest trends and avoid being characterised as outdated and old-fashioned, the term “smart” port is excessively used by professionals, policy-makers and academics. But in reality, what does having a “smart” port really mean – other than admittedly profiling a new buzzword?

The term “smart” has always made me feel really uncomfortable. Maybe because one of my favorite quotes from Einstein goes something like this: “Everybody is a genius. But if you judge a fish by its ability to climb a tree, it will live its whole life believing that it is stupid”. In this sense, maybe it’s because if we characterise some ports as “smart” then we imply that the others are “dumb” –and I cannot think of anyone that would like to characterise any port as “dumb”.

Seeking for inspiration, I turn my attention to people. Billy Bean is definitely a smart man. He is the General Manager of the baseball team Oakland Athletics and if the name rings a bell then it’s because either you are a fan of baseball, or you’ve read Michael Lewis’ book “Moneyball” (now a major motion picture). What makes Billy a smart man? In 2006 his team ranked 24th of 30 major league teams in player salaries but had the 5th-best regular-season record. To put it in perspective, in 2006 Oakland Athletics won the same number of games as the New York Yankees. Oakland paid $260k while New York paid $1,4m per win! How did he do it? With the intensive use of data analytics and statistics, Billy Beane and his assistant Paul DePodesta managed to get things down to one number and found value where no-one else could see.

Billy Bean’s accomplishments should be a guide for any executive not only in the port but also in the shipping and the wider freight supply chain sector. “Smart” is an approach built on three basic principles. Firstly, it is about doing more with less; secondly, it is about finding value where others cannot see; and lastly, it is about keeping things simple but not simplistic. In a nutshell, it’s about moving away from decisions made by gut feelings to adopting a fact-based, data-driven decision-making culture. The latter is not another “futuristic, good-to-have, abstract notion”. On the contrary, nowadays developing essential data-driven tools to support both strategic planning and daily operations is remarkably easy since Information Technology enables us to compile, select, organise and analyse data like never before.

Today, we have the essential tools and technology to move away from generalised metrics than are often misleading and may lead to inefficient practices. Working with averages, spreadsheets and simple graphs becomes a thing of the past for many industries. Shipping and ports is not – and should not be – an exception. Data and IT enable us to develop analytics, insights and predictions to the lowest level of analysis possible. Wouldn’t it be great if we could analyse why this particular shipper imported 10 TEUs that particular day of the year with a specific shipping line and not with another one? And wouldn’t it be even better if we could predict the very next time that this shipper was about to import another 10 TEUs? Or even being able to detect that a major delay and its causes will likely take place within the next 12 hours?  Sounds futuristic right? Well, it’s not – it’s already happening.

Machine learning refers to algorithms and software that learn from past data to predict future behaviour. It uses data-driven models and analytics to understand and fuse vastly increased data, and to better predict demand (for an outstanding visual introduction to Machine Learning visit:

Optimisation refers to scientific methods that model any organisation or system to obtain optimal outcomes. Advances in the last 10 years mean we can optimise multivariable, large-scale and complex systems, fixing bottlenecks and unlocking under-used capacity.

Both Machine Learning and Optimisation are extremely powerful techniques when applied properly and on a stand-alone basis. Combining both, leads to more powerful and useful outcomes. Machine Learning produces reliable predictive analytics and feeds optimisation. In turn, optimisation comes up with optimal solutions under given thresholds of uncertainty (anticipating that forecasting models cannot be 100% accurate). This is a very dynamic process, modeling increases in fidelity (and thus reliability and efficiency) as more targeted data become available.

Machine Learning and Optimisation will revolutionise the ports, maritime and freight supply chains in the next few years. And I cannot think of any “smart” solutions without having data, machine learning and optimisation (undeniably along with other great emerging trends like Internet of Things ) at their core.

Finally the greater advantage of IT applications is that they do not require significant capital investments (unlike cranes, dredging, berths etc.) but rely solely on data in an effort to maximise efficiency and find value where humans miss it. For example, using a “super computer” today costs not more than a few dollars per day while storage of data can be free or less than a dollar for several terabytes. With such an attractive cost to benefit ratio I cannot think of any good excuses for not giving it a crack.

Billy Bean had limited budget to compete in an uneven, fairly unpredictable and complex environment but succeeded with the use of data and a few algorithms. I do not see how “smart” ports, shipping and freight supply chains can be a reality soon without putting some trust on people that are like Billy.

For feedback, questions etc contact me at:


About Author'

Leave A Reply

3 + 2 =