Cubic Adds Big Data and Analytics Subsidiary to Improve Transit Ops
After spending almost $21 million last year to acquire NextBus, a Bay Area business that uses GPS data and wireless networks to predict when the next bus will arrive at any given bus stop, San Diego’s Cubic (NYSE: CUB) has been looking for other ways to breathe new life into its transportation systems business.
Today Cubic Transportation Systems is announcing the formation of Urban Insights Associates, a new subsidiary that uses big data and predictive analytics software to help public transportation agencies improve the operations of their transit networks. As part of the business, the company also has formed a consulting and services practice to help transportation planners optimize transit operations and minimize delays.
The underlying idea is to provide data-driven insights about traffic, routes, and transit operations. Cubic launched the subsidairy, based in Washington D.C., with just a handful of employees and more than a dozen technical consultants.
“We have this incredibly rich knowledge of how well things perform through the transportation world, but transportation agencies don’t make use of it,” said Matthew Cole, the executive vice president of Cubic Transportation Systems who is overseeing the company’s diversification strategy. “The data resides in their systems, but they don’t have adequate tools to bring the data together, and they don’t have the people.”
Cubic Transportation Systems’ core business makes and installs fare-ticketing equipment for mass transit systems in Miami, New York City, Atlanta, London, San Francisco, Los Angeles and other cities. The company also provides information and services that includes on-site management, operations support, business support, and other services. Cubic says more than 50 million travelers interact with its fare-card readers every day, and the company processes about $20 billion in fare revenue per year for all customers.
Cole said Urban Insights already has begun aggregating data from disparate organizations, so its analytics services “are not constrained by Cubic and what or where it is or isn’t.”
For example, the San Diego Metropolitan Transit System (MTS) is using Urban Insights to generate holistic data about commuter trips that mix segments of bus, light rail, and trolley systems. Urban Insights fuses data from five different sources, uses analytics to model trips, and identifies ways the MTS can improve its services.
Customers wanted to combine the data that Cubic’s management systems generate with many other data bases they collect to derive insights to improve their operations, Cole said. “This is something that was interesting to them,” he said. “They’ll have a better understanding now of how commuters and travelers are using the transit system.”
In the United Kingdom, “the road network is such a critical thing that closing one lane for maintenance can cause chaos,” Cole said. By using Urban Insights, transportation planners in London can run computerized simulations to model what might happen if different detours and lane closures were used to minimize the impact on traffic.
“The first area of innovation for us was in finding a way to manage very large volumes of detailed data, and to find a way to integrate them without losing any detail in the process,” said Wade Rosado, Urban Analytics’ director of analytics. To address that challenge, Rosado said Urban Insights developed a distributed data management and processing system using Apache Hadoop, an open-source software system for large-scale processing of unstructured data from a variety of sources.
“We’re willing to take disparate data sets that were never really meant to be spliced together in any meaningful way, and making it possible to use analytic software,” Rosado said.
Urban Insights also could integrate social media and other types of data to evaluate consumer sentiment, Rosado said. The data itself would remain in systems maintained and controlled by customers, but Urban Insights would “de-identify” information that might be used to identify individual users and their personal habits and preferences, Rosado said.
“I look at it like birds flocking,” Rosado said. “We’re not interested in what any one particular bird is doing.”