ORACLE TEAM USA relied on vast experience, a bold strategy, and the sheer will of a determined captain and crew to record a historic and dramatic come-from-behind victory in the 34th America’s Cup regatta. The team had another powerful ingredient for success—big data proficiency—which provided unprecedented insight into nearly every aspect of the competition and, ultimately, delivered a decisive advantage.
Trailing eight-to-one in the first-to-win-nine series, ORACLE TEAM USA battled back to win eight consecutive races—a stunning feat that made sports history. The crew sailed to success on a wave of big data supported by Oracle solutions, including Oracle Exadata Database Machine, Oracle Database, Java, and Oracle Application Express.
ORACLE TEAM USA took high tech to the extreme. Its use of Oracle solutions in the 34th America’s Cup was a groundbreaking illustration of how engineered systems, analytics, and an agile and flexible programming platform can give a captain, crew, designers, and an entire support network the knowledge to focus on their mission and win handily on a global stage.
“We engineered a high-performance vessel built for extreme sailing. On the water, the sky was the limit for this team,” said Asim Khan, director of information systems, ORACLE TEAM USA.
Data was front and center in the 34th America’s Cup—beginning with the rules-making process. In contrast to past regattas, the 2013 contest pursued an open data policy. All contenders had to provide data to the race committee so that teams had the opportunity to learn from each other.
Oracle solutions, including Exadata Database Machine and Java, factored heavily into the intense and highly iterative design, build, test, and redesign cycle.
In designing America’s Cup vessels, teams must consider thousands of component combinations. ORACLE TEAM USA used Oracle Exadata Database Machine and Java Macro Scripting to rapidly create and run 3-D simulations of how each candidate would perform in terms of speed, flow efficiency, and boat handling. The simulations required many gigabytes of data related to load and resistance, as well as historical current and wind conditions from the San Francisco Bay.
“The faster you can complete each design, build, test, and adapt cycle, the greater your advantage as teams can see upwards of a 30% performance improvement between testing and race time,” Khan said. “Oracle Exadata, and its extreme performance, enabled us to accelerate the cycle like never before,” Khan said.
The amount of data collected by the catamaran when underway was nothing short of amazing. The ORACLE TEAM USA boat was equipped with more than 300 sensors that collected real-time performance data—spanning 3,000 variables, running 10x a second—which was transmitted to an onboard server. For example, sensors captured resistance, mast strain, wing performance, and the impact of each adjustment. In addition to sensor data, the operation captured more than 200 GB of video daily, which was also fed into the analysis funnel.
The team sailed daily with a performance chase boat containing an Oracle Database that served as a real-time analytics hub. A four-person performance team configured a light feed of about 150 key sensor parameters transmitted in real time from the ORACLE TEAM USA boat to the on-shore Oracle Exadata Database Machine.
The performance team ran analyses geared toward optimizing boat performance, and fed that information to ORACLE TEAM USA sailors. One team member analyzed data from the sails and wing; another looked for data trends. A systems tech monitored the system itself, and one team member looked at the data from a sailor’s point of view. The database on the performance chase boat was synchronized with the on-shore system at the end of each training run.
“The combination of Oracle Exadata and the 5 Ghz Wi-Fi link to the boat, enabled us to move large amounts of data rapidly and easily," Khan said. “We saw dramatic performance improvements with Oracle Exadata, which runs an optimized version of Oracle Database that's been tuned for speed. “When you look at CPU-intensive tasks, there is roughly a 10x speed improvement, while I/O-intensive tasks have improved by roughly 20%.”
Sailors on the ORACLE TEAM USA boat benefited greatly from extensive real-time information during training runs. For example, they used real-time information delivered via Java-based applications to make sure they did not exceed load limits as they sailed.
Sailors wore ruggedized PDAs and received a real-time customized feed of information to help improve performance. There were also numerous tablets in fixed places on the boat that displayed more general information, such as wind speed. The PDAs and tablets on board were programmed in Java, a lightweight code that lets the team run 30 wireless feeds without slowing the wireless network.
“Since we drive the boat on numbers, it's a big problem to have information that is even seconds late due to connection problems, particularly as the boat now flies on thin foils. The foiling requires real-time information that needs to be really accurate," Khan said.
The IT team used Oracle Application Express to simplify information access via Web-based portals and mobile applications. One application streamlined quality control on the data sets that the team generated on up to 70 performance tests a day.
In addition, ORACLE TEAM USA used an Oracle Application Express-based mobile app that automated a 250-item checklist necessary to prep the boat for sailing, saving time and reducing errors.
San Francisco Bay presents a complex matrix of tides, currents, wind, and weather. Analyzing the complicated meteorological interplay between the water and surrounding land mass was a computationally intense process, which Oracle Exadata Database Machine handily cut down to size.
The support team pulled weather and tide data from publicly available sources, such as the National Oceanic and Atmospheric Administration into Oracle Database, and added information from the team's meteorologist, who created locale-specific models, based on the sensor-derived historical data stored on the team's Oracle Exadata Database Machine. From that, Khan built grid files—matrices of wind, tide, current, and weather data—and transmitted them to the performance team. The team used the data to calculate the most efficient path to navigate a course, as well as to measure the predictive data's accuracy by comparing it to real-world conditions.
Oracle Exadata enabled the team to get better results in less time. "Because it runs faster, Oracle Exadata enabled us to conduct four-dimensional data simulation, which allowed us to initialize the model runs with real observations," Khan said. "It worked well and enabled us to get the most out of the forecasted conditions. The boat had different configurations depending on wind speed forecasts, but it takes time to make those changes, so you had to make the call in advance—usually 8 to 10 hours ahead of time, which is not easy in a just-in-time construction mode. The models gave the team the added assurance—in the midst of a battle against time—that those changes would deliver a competitive advantage during a race.
This insight and agility served the team particularly well after an ORACLE TEAM USA vessel capsized during a training run. Through extensive use of technology, such as data and video analytics, the team repaired and significantly improved the boat. Barely a week after relaunching its vessel in early February 2013, the team lined up with fellow cup-competitor Artemis Racing in a training session on the bay.
Like every America's Cup team, ORACLE TEAM USA analyzed competitors' performance and racing tactics in search of a competitive advantage, and technology played a key role in its success. For starters, the team sent an observation boat to follow the training runs of competitors.
One task was to video the proceedings, which ORACLE TEAM USA used to analyze how its competitors handled the onboard logistics of maneuvers, such as tacking and jibing.
They also turned to the past for insight, pulling data from the earlier AC45 World Series races, which featured smaller versions of the America's Cup boats. "We loaded that data into our database and then ran a tack-and-jibe analysis of all of the other skippers to build up some sort of matrix on what they were doing," Khan said.
Khan and his team were keenly focused on getting critical information to the sailors as quickly as possible after they came off the water. "When they come back, they want to look at the numbers and get some objective sense of whether things were running well or not. We would look for standout variables in a certain time period to validate whatever they felt out on the water," he said.
Legacy systems took 30 to 40 minutes to collect, import, and run the reports—a long wait for sailing team members who had already put in a full day. “It is important to get them that data while the sail is still fresh in their memories. If you wait until the next day, they've lost a lot of information that was top of mind. Getting that time down from 40 minutes to 10 minutes was critical, and that's where Oracle Exadata made a huge difference," Khan said.
One tool that was particularly helpful for on-shore analysis was Race Cutter, an application that pulled sensor data from the Exadata Database with added metadata markers that synchronized video, photos, and audio streams with the raw numbers. The crew was able to click on specific moments and view sensor data from that time stamp, listen to coaching, and review a single testing sequence—continually improving the craft and sailing precision.
The 34th America’s Cup was memorable because ORACLE TEAM USA won in dramatic fashion.
It was significant because technology had such a profound role in helping the American team to win. The dynamic use of technology—from end to end—was a groundbreaking evolution in the 162-year-old international race and will likely change the competitive landscape of the regatta for years to come.
Oracle’s technology solutions supported data collection, real-time analytics, performance sailing technology, historical analysis, and database performance, giving ORACLE TEAM USA an unprecedented technical advantage.
ORACLE TEAM USA didn’t win the America’s Cup just because they sailed faster. They won because they sailed smarter.
Oracle solutions clearly supported the extreme demands of ORACLE TEAM USA in the 34th America’s Cup. As the team’s demand for real-time analytics, data management, and a robust database solution become more pressing, only Oracle could provide innovative solutions to ensure the success of the American team on a global stage.
“Oracle solutions gave ORACLE TEAM USA an obvious advantage over our competitors, and the outcome of the 34th America’s Cup validates that decision,” Khan said.