Ever wonder how all that up-to-the-minute traffic information gets to the mobile apps on your phone or the GPS on your dashboard? Between point A and point B lie billions of data points. Companies like Kirkland, Wash.-based INRIX Inc. sift through and make sense of all that data so drivers in 32 countries can get where they're going faster.
Seven years ago, INRIX was founded with a mission of changing the way the world looks at and deals with traffic. Using a crowdsourcing-based model, it tapped into mobile devices and dashboards to get real-time information. Once upon a time (and in some cases still), this data was gathered by sensors placed by departments of transportation and traffic helicopters. The company's SmartDriver Network aggregates traffic data from nearly 100 million private and commercial sensor-carrying vehicles globally to provide real-time, historical and predictive traffic information.
For INRIX, analyzing big data -- and doing it quickly -- is its stock in trade. Most companies struggle with deciding which data points are keepers, but for INRIX essentially all the data coming in is useful. The real issue is getting it back out to customers in a quick, meaningful way, said Senior Director of Product Management Ken Kranseler.
Keeping 'score' with data analysis tools
As a service to its customers -- as well as a marketing tool to would-be customers -- five years ago INRIX created a Traffic Scorecard. Initially, the annual scorecard aggregated a year's worth of U.S. traffic information to show year-over-year trends. INRIX customers, transportation reporters and departments of transportation were keen to have the information. But for INRIX, Kranseler said, there was little bang for the buck.
We figured out that with the flexibility of the tool and the way it analyzed the massive amount of data we were throwing at it, we could accelerate the process we had.
senior director of product management, INRIX
Compiling the information required pulling a data scientist away from their regular job for three months. Essentially, a mass of SQL data was downloaded into spreadsheets and crunched down into digestible bites. The end result, Kranseler said, was physical booklets and some canned Excel spreadsheets that users could download from INRIX's website. Users could see what the most congested cities were and how that changed from previous years. Kranseler and his team looked at the static end result and decided a yearly report didn't meet the needs of 21st century consumers.
"We wanted to make a much more usable report that we could update on a monthly basis and take just a few hours to do," Kranseler said. "So we were looking for a tool that that would allow us to reach into this vast database that we have, summarize it [and] report it in a dynamic way."
Several data analysis tools were considered. There was even an exploration of whether the team could better leverage Excel or SQL to meet their needs. Seattle-based Tableau Software has an office in Kirkland, and being local to INRIX put it on the company's radar. The ability to crunch the numbers and present them in a visually understandable way on its website, using the same tool, made Tableau a compelling choice.
"We figured out that with the flexibility of the tool and the way it analyzed the massive amount of data we were throwing at it, we could accelerate the process we had and create these traffic scorecards all around the world," Kranseler said. "Now when the press or a department of transportation wants to see what we have to offer [or] why the information is useful to them, they can go to our website and pick the city, month, year they want and the data flows from there."
Ease of use was a huge selling point. Using Tableau Public Premium, Kranseler's team set up a portion of the INRIX website with a header for the scorecard information, and then "snapped in" the data visualizations.
"We're now able to dynamically update it on a monthly basis; it takes about three hours to get the SQL queries done and loaded, [and] updated all in five different languages," Kranseler said.
It can sometimes be a challenge to show the business the value of a technology tool, but the proof here was the new, dynamic scorecard itself.
Read more about how companies are taking on big data analysis
Big data management? More like big decisions management
How CIOs can avoid big chaos through a BI data steward
Hadoop offers big data analytics on the cheap if you know how
"The scorecard was a nice thing to do, but it didn't necessarily add direct value for a specific customer or dealer or revenue opportunity, so it was always hard to find revenue sources for it," Kranseler said. "As we grew and added more insights that we wanted to provide from our data, [we] either weren't going to continue or it was going to take a lot more resources to go at it, so we said either we stop doing this or find a more scalable way."
The response has been overwhelmingly positive, Kranseler said. Departments of transportation and other public sector groups are thrilled they now have the information they lacked the funding resources to gather. Urban planners and journalists are also fans. It also helped win INRIX new business in Europe with the U.K. Highways Agency. While the agency was considering INRIX as a private partner to provide source data for traveler information services, INRIX published its first European Traffic Scorecard with U.K. data.
It was a core element of what the agency was looking for, said INRIX Community Relations Manager Jim Bak. "It helped boost our credibility and made their decision process move more quickly and won us a seven-year multimillion-dollar deal, so it's really been a tool that's helped drive the business forward from a marketing point of view as well."
Dealing with data in new ways
Having the tool also has allowed data scientists at INRIX to analyze data in ways they hadn't been able to efficiently accomplish before, Kranseler said. One example of this is determining data quality. In the past, if customers reported a problem -- say the congestion reported on a segment of road was inaccurate -- looking into that complaint was "like looking for a needle in a haystack of data" to determine whether the problem was endemic or systemic, Kranseler said.
Using Tableau, the INRIX science team is able to visualize what's happening on a specific stretch of road and how it varies over time. For example, if there were data quality problems in Munich during November, the team can conduct a root cause analysis by drilling down to see what was happening, when and why on a particular segment of road.
"There's a standing body that looks at these problems, and now we have a platform that we can go present from to be able to have a common dialog about how these problems come up, where we can diagnose and fix them, and put the right solutions in place to make sure our quality keeps getting better over time," he said.
Let us know what you think about the story; email Karen Goulart, features writer. For more stories on big data analysis, check out our Business Intelligence for the Midmarket story archive. For midmarket IT news and updates throughout the week, follow us on Twitter @ciomidmarket.