News Article 21/10/2011 A
Cloud Formation (theBaltic Autumn 2011, p.61-63

What’s new and what’s relevant in IT for shipping in the next new years? Dimitris Lyras, founder of global software company Ulysses gives his view.


The mainstream IT industry is paying a lot of attention to text analytics. This means searches that detect discussions that require the attention of the company or enable the classification of typical documents. Things like customer complaints on Clogs, customer complaints that have not been answered or indexing of legal documents. An indication of this is that Hewlett Packard bought a leader in the field of text analytics, a company called Autonomy or for over eight times its revenue, US$10 billion.
Analytics is relevant to any enterprise, including shipping. Everyone wants to keep track of emails, and content and search tools don't help much. There have even been a few users of Autonomy in our industry.
Business Intelligence (BI) is also still a vast and popular development in the mainstream software industry. Originally, it was a study of numbers to arrive at conclusions that were not obvious by general observation. A very popular use of BI was, and still is, statistics on customer trends under a variety of circumstances that could lead to a better product related sales or distribution strategy, for example. Today, the critical element is getting the BI projects deployed much quicker than before, which in turn is about making new reports and graphs more quickly from vast amounts of currently generated data, without slowing down the company servers and databases.
Social media is still getting a lot of attention in main stream software. We in shipping have our share of social media — discussion groups, Twitter and blogs — as do other industries and the value of such places to discuss things has some benefit over writing to your friends via email or some such alternative. The importance and power of grouping people with interests in common cannot be overestimated.
The cloud is another major development in the mainstream software industry. The cloud is really about mobile devices having protocols to engage with applications anywhere in the world without the user having to log on and off a website ad infinitum, while moving around in a car for example.
These are not the only main stream popular developments or the newest ones, but they represent where the big money is going.
So how does all this mainstream technology expenditure apply to shipping?

Text analytics
All industries can make use of search to categorise and find textual information. We all face this when we have to look through emails of more than a week to find an internal discussion.
However, the concept of categorising information is perhaps more important than anything; even more important than the issue of finding information easily, ferreting out emails or say finding and reusing wording for legal clauses.
So text analytics gets involved with categorising information, which is even more important to an enterprise that search itself.
Categorising information is actually what we do all our lives and what makes a person with experience able to make good decisions. We categorise past experiences in order to assess a current decision. Even a dog categorises experience about his master in order to predict his master's behaviour. Categorising information is also what we do when we co-ordinate to solve a problem. To co-ordinate we have to know what relevance the problem has to each stakeholder in a company so as to get them all to contribute in time to minimise risk.
Categorisation is also what can make statistically derived data useful. Those of us who have worked with accident statistics know that it is notoriously difficult to agree and assign categories to incidents. Do you assign the accidents to their root causes, do you assign the accident to its effect, for example fire or explosion, do you assign it to their immediate causes etc? More importantly, do you assign context to each accident so that you don't confuse a fire during repairs in a yard with a fire during normal ship operation.
So the fact that analytics is big business means that soon categorising information will be big business, which will change everything in how we manage information.
So watch this space!

Business Intelligence
Business Intelligence is most useful when there are hundreds of thousands of pieces of information gathering every day and telling people about their business. In shipping, internal information useful for business intelligence is much more coarse-grained and much less frequent and, considering the small volume of information, it is staggeringly varied. There are at least 1,000 different events and associated data-points, ranging from a deck crane that is overheating to a fuel barge that is running late that could each potentially change the revenue of a ship’s voyage by say $10,000. However, each of these events has frequency that is so low that you could hardly ever make statistics out of it. In short, shipping has too many processes and too few repetitions of each process to benefit from Business Intelligence the way retail industries can, for example. Sure, we like to see graphs of daily fuel consumption, say, but this hardly requires cutting edge Business Intelligence.
More importantly, BI is in need of categorisation as well. Even daily fuel consumption needs to be categorised by ships draft, absolute speed, through the water speed, wind, currents, sea condition etc. Breaking it into the right categories and gathering enough information about each category is much harder than making graphs and preventing your data server from grinding to a halt from the load. For example, accurate through water speed is very difficult to achieve from the hardware, and once you reel all the consumption information into the categories provided, it is difficult to gather the millions of data points in each category so as to justify state-of-the-art BI. So our BI problems in the maritime industry are not the same as those in other industries where this technology is most popular. We need to make statistics but we need to break them down into meaningful categories, and non external technology is going to do that for us. It's a business issue not a technology issue. That, for example, is the reason Intemanager has worked so hard to provide KPIs for quality.

Social media
Every industry and every person needs a group of people to discuss different interests. We also all need to network so as to leverage each other's knowledge and efforts, meet the right people etc. In shipping we have no less reason to do so than in any other industry. There are many obvious examples of the value of social media.
However, the technology enablers will develop in aspects of social media that gain widespread revenue for the software companies involved. One size fits all will come first, long before the specialisation needed to really serve an industry. LinkedIn was sold for $6 billion or so, and we are seeing Hollywood movies about the success of Facebook. These are one size fits all systems, the kind that make the most money for the developers. But it is not where the real value lies. Despite all the success of Facebook and LinkedIn, how come we are asked questions such as "Like" "Don't Like" to categorise a new posting? When did an unqualified "like" or "don't like" ever help us find the right article to read? How often does a simple discussion group, as we often see in social media, provide us with discussions we really care about?
So it will be a while before we see the real value of social media but there is no doubt it will provide more and more value as the categorisation and qualification of the social groups and discussions becomes more sophisticated.
So we mentioned categorisation again. Of course the future lies in categorisation. Would we not all find social media more useful if before reading something from the stacks of information available every morning, we knew what the point of the discussion was and how people like us might find it useful? Let's not forget that the mainstream media, which is having attention diverted from it by social media, has a publication name that is at risk when publishing an article, and journalists with experience, knowledge and expertise. All this is qualification and categorisation that needs to be put back into the social media.

The cloud
What does shipping really need in the next few years from the technology available?