Ellison out to steer Oracle yacht in the right direction

NET RESULTS: LOTS OF technology companies like to show how they “eat their own dog food” – use their own products and services…

NET RESULTS:LOTS OF technology companies like to show how they "eat their own dog food" – use their own products and services internally in the company. Doing so offers up statistics and case studies of how their products can be used by other organisations.

But the penchant of a certain company’s billionaire founder for racing the world’s fastest sailboats has meant it can carry off this eating business with unusual flair.

Yes, of course, we’re speaking of none other than Oracle’s Larry Ellison, whose Oracle BMW team in 2010 won the most recent staging of the America’s Cup, a sailing race first held in 1851 that rolls around once every three years and which is preceded by many racing regattas.

Oracle, busy trying to create awareness about and sell into the “Big Data” market, doesn’t have to look far for a sexy example to show customers how gathering in and analysing patterns in vast amounts of data can give a competitive boost.

READ MORE

In Oracle’s case, the data streams are coming from sensors all over its sleek racing catamarans. Those data are analysed and then directly influence both boat design (in the form of the final yacht for the actual cup race) and sailing technique.

The person whose job it has been for many years to analyse sailing data to increase performance for Oracle is Ian Burns, who, with the next race in 2013 in San Francisco, is involved in his ninth America’s Cup.

I first heard him talk about the technology of sailing back in 2005 in Valencia in Spain, during the races leading up to the 2007 Cup final. Back then, Oracle was putting 120 sensors on its single-hull, carbon fibre, 17-crew racing yachts and using its existing database analytics to try to make sense of the information generated from them.

In the last America’s Cup, Burns was dealing with 250 sensors pumping out 2,000 parameters of data 10 times a second. This time around, he’s got 500 sensors on five-crew catamarans, which can churn out 5,000 data variables 10 times a second.

Some 50 or 60 of those sensors are just for windspeed and angle, he says.

For the first time, Burns will have some serious data-crunching tools, in the form of a new range of Oracle hardware and software, to work through those data streams.

“We dabbled with these tools during the last America’s Cup [while they were being developed] and now we are getting going into this more heavily,” he says. “It’s actually very similar to the business situation – we’ve got a whole lot of data and often, we don’t know where it’s going.”

Like businesses that are taking in terabytes of information off sensors, websites, social media streams, inhouse networks and other sources, it can be difficult to tell which data parameters are important and which aren’t when they run into the thousands, he says.

Once Oracle Racing started using the latest tools, “we added more sensors and began to find other relationships”.

For example, “there were big differences in performance in moving the [steering] wheel in certain patterns that we hadn’t seen before. We’ve also proven two or three things we didn’t really know before.”

Before they had the sensors, a sailing team could test just one parameter a week by focusing on that one element during practice and races. Now, they can look at hundreds or thousands a day if they wish.

They really have only just begun, with races running at the moment around Europe – they were in Plymouth late last year, in Naples for a couple of weeks over Easter, now they move on to Venice and then over to the US, to Newport, Rhode Island.

Over the next couple of months Burns will be poring over sensor data returning from these events as the team begins to plan its final two boat designs. The first will launch in July and the second in February of next year. They’ll compare performances and only one will end up in the America’s Cup in 2013.

“We’ll get all that data together and look for answers question by question,” he says.

The last winning Oracle yacht improved its performance 70 per cent from its launch to race day, he says. “That shows the significance of performance analytics.”

It’s all a bit of a contrast though to when he started in the America’s Cup. “We had four sensors – the compass, boat speed, wind speed and the angle of the rudder.”

Of course great sailing performance is not just about the sensors in the computers, he says. “We capture the learnings from the wind and water and bring that back to the sailing team. We can tell when a boat is going well, but we just can’t tell why. A great sailor will distill those answers.

“The human element [in sailing] remains the same, but now we have some great tools and can accelerate the learning much, much more.”