Big Data and analytics must be used and grounded in a way that reflects life at sea – and this means preparing for the unexpected, argues Huw Davies, Principal of Meteorology at StratumFive.
If you were to pick one of the most overused words in shipping in 2017, ‘digitalisation’ would likely be near the top of the list. It is now almost mandatory for maritime companies of all types to put ‘digital transformation’ at the heart of their strategies for growth. However, could most people in our industry name a tangible example of how Big Data and digital solutions make a tangible difference to their operations? Too often, the value of Big Data remains short-lived, vague, and perpetually out of reach. The topic is seen as the domain of futurists and marketers, rather than those involved in the day-to-day operations of shipping.
At StratumFive, we see Big Data as something that can help seafarers, owners and operators make better decisions, and face the unexpected with confidence thanks to the power of predictive models and machine learning. However, the onus is on companies like us to demonstrate that Big Data can deliver real results that tackle real issues.
One such example is this interactive heatmap, [see image below] which highlights the relative risks of piracy in different areas. Based on this information, a captain can decide where to accelerate through potentially risky areas, and where it will be safer to slow down and save fuel. This map is based on machine learning; a process that examines the relationships between factors, and works out which are the most important. After determining which cause and effect relationships are the strongest, we can then build an accurate predictive model.
In this instance, many of the influential factors are as expected; wind speed, direction and wave height and swell direction are the most important, as well as light levels. However, the day of the week also plays a part.
On the face of it, this may seem illogical; is a risk of piracy really worse on a given day of the week?
The answer, it turns out, is yes. In Somalia, Fridays are days of prayer. Pirates, it turns out, can be divided into two groups. Less experienced, opportunistic, ’part time’ pirates, and hardened ‘professional’ pirates. The former group will observe their holy days, while the latter will venture out regardless. Because of this, if a pirate attack occurs on a Friday, it is more likely to result in a hijacking. The understanding derived from this analysis will shortly be incorporated to the security applications within OTiS, our vessel tracking software – allowing users to directly benefit from this research.
This exemplifies the approach that is necessary when it comes to Big Data. Rather than looking at individual aspects of the voyage, we need to take as many factors as possible into account when we develop our tools and analytics. Factors which may not seem initially obvious – such as the day of the week – can be highlighted with tools that work against our unconscious biases and assumptions.
This underscores how we think about Big Data solutions at StratumFive. Since 2009, we have providing insight-driven technology services, including advanced voyage monitoring, next-generation ship security alert systems, security intelligence, fleet tracking and meteorology.
We believe that this is the future, and the way that shipping will fully realise the benefits that powerful new analytics tools can offer. Currently, we see a ‘goldrush’ mentality emerging in shipping, where many are tempted to go after the latest ‘shiny new object’. However, many of these are built in a very narrow fashion – focusing solely on one element of the voyage. These solutions are often the result of teams working in silos, and failing to share the relevant knowledge. For example, we know that a system for trim optimisation can be very valuable. However, a poorly thought out course change can render any benefits of trim optimisation useless. No matter how powerful, any system that doesn’t take a total view of the voyage as a whole is not fully meeting the needs of the industry.
This is what we are determined to avoid as we move forward and develop new solutions, building on our existing legacy of trusted safety, security, weather and tracking systems. We believe that the best way to meet the industry’s needs is to go beyond offering individual products and pieces of software, and instead offer a platform for total voyage intelligence. The principle behind this is a platform that is easy to use and access, that promotes knowledge sharing between teams. At the same time it must be flexible, searchable, and powerful enough to bring together varied datasets from multiple sources to create the deep insights highlighted above. The more data the platform handles, the more powerful and useful it becomes – so we want to make it as easy as possible to collaborate with multiple and varied different partner organisations.
Ultimately, while digitalisation may be new to shipping, the mindset needed to make the most of the opportunity has been present in the industry for a very long time indeed. This industry is used to considering the whole picture, from the beginning of a voyage and beyond a cargo’s arrival into port. The data solutions that we employ in the future must do the same, and must be developed in tandem with the industry. By adopting a pragmatic approach derived from knowledge and expertise stemming from real life experience at sea, shipping can truly put Big Data to work for its benefit, ensuring that the industry is equipped to cope comprehensively with the ‘adventure’ that any voyage constitutes.
By Huw Davies, Principal of Meteorology at StratumFive
Huw has a very strong business and maritime background. Joining the Royal Navy from University, he served in a number of roles at sea and ashore in UK, US and Europe before joining the Senior Leadership Team at Rolls-Royce. He has also worked extensively with SMEs having founded and assisted technology start ups. He has Masters degrees in Business, Meteorology and Oceanography and International Affairs and is a technical author published by the Nautical Institute.