In Data Science, Data Management Is King
- By Winston Thomas
- May 11, 2020
Burned rubber, screaming engines, sparks from skid plates and wafts of high-octane emission define the excitement that surrounds F1 races. But beneath the bravado of the race drivers (and their caustic attitudes) and the marvel of engineering, there is a second race that occurs in the ether. It is one that is fought with data science.
A tour of the Mercedes-AMG Petronas Formula One garage in Singapore last year showed how much F1 depended on data science. A similar tour two decades ago showed engineers poring through numbers, comparing historical charts and chatting with the drivers. Today, they all talk the same analytical lingo — and with Mercedes-AMG Petronas, this conversation is powered by TIBCO Software.
Data rush
Around 300 sensors on an F1 car flood today’s races with valuable data. It means that F1 teams are watching live data as the F1 cars outmaneuver each other at speeds topping 340 km/h.
Where the winners are heralded is how this data gets interpreted, collated and turned into insights at blinding speed for team principals like Toto Wolff. It makes data management, rather than pure data analytics, crucial.
Strong data management allows teams like those in F1 to do three things: properly prepare for different race conditions or scenarios; formulate the best strategies on the fly based on how the race is progressing; deal with unexpected issues and emergencies as they occur.
To do all three, the team needs to address two major challenges: immense data volumes and data coming at the team at warp speeds. This is where TIBCO Spotfire and TIBCO Data Science come into play.
Predictive algorithms help the F1 team to analyze past races and circuits to understand what worked and what did not. These offer valuable insights as the team prepares for the new circuit — and every circuit is unique and has its own nuances. Meanwhile, performance trends can help the team focus on key metrics that matter and decide on split decisions.
Crashing waves
While data analytics are not meant to replace human decision making, TIBCO Software’s work with Mercedes-AMG Petronas Formula One offers a blueprint on how it can be done at high-speeds. It also opens new doors to innovation.
Consider application testing. You can consider F1 as a high-octane proof of concept, with the racing car as a perpetual MVP. The technologies, innovation and expertise are cutting edge, with many still experimental. The data science helps the teams understand which innovation worked, and which failed.
This “fail fast and often” mantra drives many of today’s startups and leading enterprise innovation teams. It offers a common language that ties the different parts of an organization together.
CargoSmart is one such company. While cargo vessels are a far cry from the zippy, lightweight F1 cars, it still requires fast data insights.
“We began our relationship with TIBCO in 2007 when we started building an event-driven architecture. It's a platform that processes a large amount of data from many different kinds of sources,” said Ralph Ho, senior manager in customer integration at CargoSmart in a published case study.
The global shipment management continued to evolve. In 2019, it added TIBCO Streaming software to correlate different types of information and help identify critical exceptions in real time, and TIBCO Flogo software to integrate a mix of data types and support innovation using various services, platforms, and IoT devices.
Now, the company is adding AI to augment decision making, especially with disruption handling. These were crucial as medical supplies became stressed and urgent during the pandemic crisis.
"We developed a vessel speed and route monitoring application that analyzes the vessel's speed and distance against a complex variety of factors. The application has helped ocean carriers reduce fuel consumption by up to 3.5% over the past two years,” added Ho.
Cashing in on experience
Data management and analytics helped Kasikorn Bank in Thailand (KBTG) to transform into a digital competitor. It used TIBCO Data Visualization to improve data delivery and put the power of information in the hand of the customers. The TIBCO Software solution helped to stabilize the performance and stability of its apps.
“A stable platform is key to everything. If the platform is unstable, the customer is not going to use your product because they don't like the experience,” said Fred Roteseree, deputy managing director at KBTG in a separate case study.
The solution saved the time and effort of moving data into a single source. Enormous data volumes in over 4,000 tables from a complex data warehouse with 35–40 data marts can now be turned into a report quickly.
"Now that we have this technology, people start to think about data differently," said Roteseree.
For example, users can view every banking product and account they have with the banks in a format that is intuitive. In comparison, traditional banks store this data often in multiple systems.
Normalizing the new normal
Strong data management backing up fast analytics will be crucial as businesses weather the pandemic headwinds. The economic landscape is shifting fast, and as companies start to get their business back to full operations, they will operate in a new normal.
Those with the right data management technologies will have the right data, insights and capabilities at their fingertips. The rest is up to those who use these insights to make decisions.
Correctly done, it can give you a podium finish — just ask Lewis Hamilton.
Photo credit: iStockphoto/FotoMaximum
Winston Thomas
Winston Thomas is the editor-in-chief of CDOTrends. He likes to piece together the weird and wondering tech puzzle for readers and identify groundbreaking business models led by tech while waiting for the singularity.