BUSINESS INTELLIGENCE: A DEEP DIVE

To me, business intelligence is taking mountains of raw data and turning it into information that’s useful. That could be aggregation of existing data. It could be leveraging that data to predict what may happen in the future. Or noticing what’s trending. One of the main things that has made Transit Miner really popular is that it exposes exceptions and anomalies in the data.

Say you’re looking at on-time performance for an agency and comparing it week over week. One day, your on-time performance is 90 percent, and you gave a thousand rides. Another day, your on-time performance was 95 percent, but we only had 800 rides. It explains the discrepancy right there.

Another example would be predicting passenger behavior. For instance, in LA, there’s 200,000 registered passengers in Orange County. There are probably several dozen that take an inordinate amount of time to load or unload. Unless a driver actually notifies dispatch or management, that would go unnoticed. But Transit Miner would identify, Hey, you got some outliers here. Perhaps it’s a problem with a specific type of equipment or access to a door. Maybe the client expects you to come in the back way, and the driver usually comes in the front way. Once an agency is aware of the issue, they can work to fix it.

Another example might be with—let’s call him Joe the Driver. Maybe the first Monday after payday, Joe the Driver always calls out. It could be something as simple as that. The agency could plan on Joe not being there that day. Another outlier example would be if a driver pulled out at eight A.M., but if you compare that to payroll data, you find that the driver clocked in at six-thirty. What was the explanation for that hour-and-a-half lag? In transit, typically, there should be a 15-minute pre-op check, or a standard amount of time that’s allocated for a driver before they actually pull out. That discrepancy presents a huge cost factor for the agency. That’s business intelligence.

It comes from a variety of places. There’s scheduling and payroll software. Transit Miner currently links to ADP. We are capable of receiving vehicle maintenance data or insurance data, to predict risk or vehicle accidents. We can review the onboard drive cam to monitor incidents. Basically, all these data sources can enable a transit manager to look at the whole picture.

Transit, in my opinion, is behind the tech curve across the board. That includes scheduling software, BI, everything. A lot of the programs that agencies call business intelligence are still using old technology, like Visual Basic, Microsoft Access, or Excel pivot tables. One of my primary goals has been to help change that so public transit starts to set standards. We have so much data that’s just begging to be put to work!

Transit Miner is integrated with Routematch. It’s available to current Routematch customers as an add-on option. But an agency doesn’t have to be using Routematch to use Transit Miner.

Transit Miner is at one minute right now, but it will be at 15 seconds with the 7.0 upgrade to Routematch. The limitation hasn’t been on Transit Miner’s end. It’s been on the scheduling software’s end. Basically, if you pull data more frequently than one minute, it can impact dispatching functionality within the underlying scheduling software because of the lack of optimization in the databases. With the Routematch 7.0 upgrade, we should be able to get real-time data pulls at 15 or 30 seconds.

Yes, I love doing that. It’s a great idea to speak with people who are using the software that you’re thinking of getting. You can ask how they made their decision, how the implementation has been, and any challenges and successes faced along the way. If an agency is interested in the software, we’ll come to you and show you how it works. You can see the app and play with the various features. In a live demo, we’ll walk you through all the functions. General Managers will often invite members of their team—like the IT or Operations Manager, or a driver—to ask questions and see how they would use the technology.

A big thing with Transit Miner is that it improves data quality. Data quality in public transportation has historically been terrible. A lot of data that’s key to quality performance metrics reporting is either missing or erroneous. Transit Miner finds the missing or erroneous data and highlights it. For a lot of our customers, when we first install Transit Miner, they find that a lot of the data they need is missing. In one case at an agency, they were missing thousands of odometer readings. It came down to drivers not hitting a button that logged it. Through driver training, that agency can solve the problem and get more accurate data.

It is absolutely. It’s just as relevant to the ten-bus operations as it is to the 800-bus operations. That’s because at a ten-bus operation, you have fewer people doing more roles. Having access to aggregated information means nobody has to run a ton of reports. It makes everyone’s job a lot easier. Using Transit Miner, a small-bus operation can get insights without having a whole department devoted to it.

You’re operating in the blind, really. You’re going off of instincts, hunches, or speculations. You’re reacting instead of making educated decisions based on facts and science.

All the intelligence that’s built into Transit Miner is centered around improving the passenger experience. We’re pointing out exceptions and anomalies in driver behavior and scheduling problems. All these things are being highlighted to the management. A lot of it is being done predictively. So, five hours from now, we’re anticipating Mrs. Jones is going to ride an hour and a half today for a three-mile trip. It points that out to dispatch, so dispatch can make scheduling adjustments to improve her ride. Right now, the software isn’t passenger-facing. But that is in the works.