Data collection and analysis – The journey has just begun
By Marsha Moore
The explosion of data available across all enterprise touchpoints means big things for the future of public transit. It used to be that historical reporting was enough. But as the amount of data being captured grows, so does the desire to improve analysis capabilities to better leverage changes in technology.
Advanced solutions that can move the agency from a reliance on purely descriptive analysis, through predictive and eventually prescriptive analysis, mean it’s now possible to use the full wealth of enterprise data to ask and answer forward-looking “what if?” type questions in real-time and, eventually, automate intelligent transit systems to implement solutions before problems exist.
Unfortunately, in the bottomless sea of rising data, it can be difficult to know what’s actually valuable, or how to turn it into business intelligence (BI) that supports the agency’s business objectives. These challenges are compounded by the complex interplay of data from different systems, including a growing variety of on-board sensors.
Getting to the point where we have metrics that reflect BI’s value to the agency’s strategic goals and demonstrating how such solutions can grow with industry and organizational shifts is an essential first step. But moving from descriptive through predictive and prescriptive analytics is complex. To envision the whole of the journey, let’s start by considering where we began.
Leaving the station
In the transit world, data analysis is traditionally thought of as something available through a set of reports. While these reports are generally suitable for back office trend analysis, front office users that require real-time data are left with a gap as the data was typically 24 to 48 hours old (i.e. you’re “looking in the rearview mirror”). Front line staff need a pervasive solution to visualize the information needed to support time-critical decisions that must be made throughout the course of each day, many of which impact bottom line costs.
To enable this solution, agencies must first identify the most significant historical data points from across the enterprise and study them for their business value (Big Data is the whole mine; Tiny Data is where the gold is). This is very much the realm of descriptive analytics: using data aggregation and mining techniques to answer: What has happened before?
On your way
Once descriptive analysis is enabled, agencies can start thinking about asking questions of their historical data to help determine how to optimize business today (a.k.a. predictive analytics, or using statistical models and forecasting techniques to understand the future and answer: What could happen?).
But, in order to answer these questions, data must first be migrated from multiple source systems and combined in a central warehouse designed to store the information for 10 or more years and still be flexible and fast enough to provide real-time answers to queries. The first logical step is to use this data to trace where problems exist by analyzing both historical and real-time data across the enterprise, thus enabling the agency to be proactive in identifying and correcting problems in real-time.
The journey continues
As the ability to leverage real-time and events-driven data (think: alerts) continues to improve, agencies cross into the realm of prescriptive analytics, which use optimization and simulation algorithms to advise on possible outcomes and answer: What should we do?
By integrating with statistical modeling platforms such as SAS and R and creating simple tools to access and visualize data, the agency will empower users across the enterprise to become proactive in solving problems and remain continuously focused on finding better ways to operate the business.
Soon, connected, integrated data from across the transit enterprise will help solve old challenges, while creating new opportunities to use business intelligence to inform decision-making and measure the impacts of those decisions.
That’s where our journey of discovery through descriptive, predictive and prescriptive analytics will lead the public transit industry. Trapeze Group’s goal in all of this is to deliver the industry’s most advanced and scalable data collection and analysis systems and ensure that we have the right solution for your agency, wherever you are in your journey.
Marsha Moore is chief technology officer at Trapeze Group. Moore has more than 30 years of IT experience and has remained a forerunner in software development, specifically in passenger transportation. She has held executive leadership roles where she drove innovation as well as developed and designed tools that increased productivity. Visit www.trapezegroup.com.