ShipInsight Trusted Information on Maritime Technology and Regulation The Quarterly Journal Summer 2015 Big Data Power & Propulsion Innovations Nor-Shipping Winners Paints & Coatings Fuel Switching Big Box Ships Fuels & Lubricants Learning from Leisure The IMO Vote VARDs New Bridge Global Xpress
36 THE BIG STORY
37 The Buzz of Big Data Every so often a new concept or idea sweeps through the world, stirs up huge interest and opportunities and then either becomes established practice or fades away. Right now shipping is being caught up in just such a whirl albeit a little later perhaps than some of the shore-based industries. Big data is big business right now, even if some are not entirely sure what it is and what it can do. In fact, there are some who would say that there are so many definitions of what constitutes big data that the term can be misleading and even off-putting. Three definitions given below are probably so vague as to leave most people unenlightened and wondering if it has a place in shipping or their own particular operation. 1. Big data is a broad term for data sets so large or complex that traditional data processing applications are inadequate. (Wikipedia) 2. Big data opportunities emerge in organizations generating a median of 300 terabytes of data a week. The most common forms of data analysed in this way are business transactions stored in relational databases, followed by documents, e-mail, sensor data, blogs, and social media. (Intel) 3. Big data is the term increasingly used to describe the process of applying serious computing power the latest in machine learning and artificial intelligence to seriously massive and often highly complex sets of information. (Microsoft). During the Maritime CIO Forum in Oslo during Nor-Shipping, KVH CEO Martin Kits van Heyningen urged maritime operators to utilise big data in order to remain competitive. Probably the most important thing for maritime managers to do is make big data a priority, he said. It s important to adopt a big data mindset, even if you don t think of yourself as a data company. Data is becoming a resource in its own right, and offers incredible possibilities for understanding every aspect of your business better. During his presentation, entitled Turning Big Data into Big Value, Van Heyningen noted that the maritime industry has been slow to adopt big data even though the industry faces many challenges for which data capture and analysis can provide answers from meeting an increasing number of maritime regulations to improving the fuel efficiency of vessels underway. The maritime industry has spent the past 20 years trying to limit the amount of data going on and off vessels, while the rest of the world has been doing the exact
38 THE BIG STORY opposite in adopting big data, he said. Computer analysis of big data goes far beyond human capacity in providing information that can make a maritime operation more efficient. For example, with real-time analysis of such data as engine monitoring, consumption rates for various fuel types, the fixed running costs of a ship, and weather data, a maritime operation can optimise a voyage for financial performance rather than just time or distance. This is not a calculation that can be done by a human, no matter how much experience with a given route the people onboard may have, said Van Heyningen. Utilising experts for remote analysis of big data can also help with preventative maintenance to avoid costly repairs. There are many people within the industry who could agree in part with Van Heyningen s views and probably an equal or even greater number that would take an opposite position. Until quite recently, communication costs and the reliability of communications have not been conducive to transmitting data between ship and shore. Nor, it must be said has the analytical software reached a state where it could provide much in the way of guidance. Furthermore, there is often a lack of knowledge about ship operations both commercial and operational that those within the industry despair at when assumptions are made about their expertise and the magnitude of savings or improvements that are possible. Proponents of big data expound the benefits pointing to studies such as one by the Massachusetts Institute of Technology which found that data-driven firms perform 5%-6% better each year. There s a growing divide between companies that use big data and those that don t, said Van Heyningen, who also believes that dramatic changes in the affordability of data analysis make this the right time for maritime operators to embrace big data. You need to be creative about the questions you want big data to answer for you. It s more important than ever to work with IT partners and satellite communications providers that can do more than just provide connectivity,
39 but can also help you solve your real-world problems, he said. DNV GL s Tor Svensen is another believer in the potential of big data and expects that it will enable the shipping industry to intensify its focus on enhancing operational efficiency. By bringing together and analysing both data from on-board monitoring systems and from external sources, a comprehensive insight is gained of voyage, engine and hull performance, he said. Voyage management based on shipboard sensors and AIS data, for example, can help to determine the optimal speed in all conditions and thereby reduce fuel bills. These views are not entirely new thinking, some pioneer operators have already begun monitoring engine performance and making use of remote diagnostics for maintenance and that is a trend that began five or more years ago and is increasing. A recent example is the agreement signed in April between Wärtsilä and Golar Management covering 13 of the operator s LNG carriers. The service Utilising experts for remote analysis of big data can also help with preventative maintenance to avoid costly repairs. agreement includes remote monitoring of 52 engines, maintenance planning, advisory services, and guaranteed availability of personnel as well as spare parts. Golar s LNG carriers and FSRU s are all equipped with Wärtsilä engines. There are good reasons for a slow take-up of remote monitoring including the fact that in shipping, many engines are not produced in great enough volumes for data to be rapidly accumulated. Even engines that are may not be operated in similar manners some may be propulsion engines, some auxiliaries and the choice of fuels and fuel treatment systems may not be identical, Right: KVH CEO Martin Kits van Heyningen.
41 Exploring the possibilities of big data is something that every shipowner or operator will have to consider. spare parts may sometimes be OEM parts and others lower quality. In addition, ships do not remain in the same hands for ever and successive owners may not be interested in maintaining the relationship with the engine maker. It is in their understanding of operational matters that proponents of new concepts such as big data often fail to understand why shipping as an industry rather than individual operators may not be bursting with enthusiasm. Despite the commonly held view of shipping being dominated by large fleet operators, the reality is that the average shipowner even those whose may have a number of ships that are registered as single ship plate companies will likely have a fleet that can be counted on the fingers of one hand. Those ships could be operated in a number of ways, such as running on own business, playing the spot market, time chartered or a combination of all three. Even some of the larger liner operators fleets will not consist entirely of owned vessels with a 50/50 mix being common and some using almost entirely chartered in tonnage. Exactly how much data an owner might generate beyond that emanating from the engine is therefore debatable. Operating ships is not all about saving fuel under all circumstances. Liner operators have schedules that they must maintain if they are to retain their customers confidence and loyalty. Even with slow steaming having stretched the span between ports on the line, the reliability and on-time arrivals have deteriorated noticeably. Weather routing can give some assistance on transocean crossings but is of little use when the time between ports at each end of the voyage is just a day or two. In the spot market, criticism of the sprint and wait strategy has become common place but this is often dictated by the cargo interests demands and should not be so quickly dismissed. A ship that is advised to take a slow voyage aimed at arrival just hours before the laycan date could easily be delayed by a mechanical problem, weather or even assisting in the rescue of boat people migrants or diverting to land a sick crewman. Failure to make the date even by an hour could result in the fixture being cancelled even if the ship has made a voyage of several days to reach the loading port. To suggest that big data can solve such problems and that experience of masters and owners is deficient is to dismiss all that underpins the shipping industry. Where big data could perhaps be of more use is to government and port operators container port operators in particular who can combine import and export data from government authorities and trade associations to better anticipate demand for services and storage space and develop infrastructure to meet anticipated needs. Whether the data can anticipate a line operator s decision to invest in new tonnage that exceeds the port facilities or whether a port in a neighbouring state is planning to poach a major customer. Exploring the possibilities of big data is something that every shipowner or operator will have to consider as being worthwhile or not. It could be prudent to ask at the very earliest opportunity what the cost of gathering and analysing the data is likely to be and to compare that with potential savings. It could be that the difference between the two could be marginal or even an additional cost if the wrong consultant is chosen to aid in the implementation.