Our Q&A pro is Routematch Chief Product Officer Sam Pandolfo. He is speaking on the subject of Next Generation Scheduling.

What is Next Generation Scheduling for transit?

Faster, more accurate, and more efficient scheduling that solves the problems of your current service delivery model and allows you to offer new on-demand services in the future.

What problems does Next Generation Scheduling solve?

Well, the big problem we’re trying to help an agency solve is that a group of people needs to get from Point A to Point B. So what is the best way to move people from Point A to Point B? For each agency in every community, there are a ton of considerations, but the basic math problem is: I have a certain number of people to move. I have a certain number of vehicles to do it. I have certain rules or constraints that I have to live within. Our technology helps the rider understand the services available to them—so that they get from Point A to Point B in the timeframes they want. And then our technology helps the agency deliver and monitor the service according to their unique parameters.

How is that different from the current scheduling engine?

Routematch’s traditional scheduling software takes all of the trips booked and optimizes the route and schedule according to each agency’s rules and regulations. Drivers will get their schedule, and they will pick up and drop off according to the manifest. But in the traditional planned paratransit services, you’ll have a 20 percent cancellation or no-show rate. That means if you have 5,000 trips on the schedule, a thousand of those trips will not happen as planned. Already, off the bat, 20 percent of your plan for the day is wrong. So how can an agency know something in real time and dynamically respond? And how can an agency perform real time, on demand trips? If I’m a paratransit rider, and I want to go to the grocery store an hour from now, can I be accommodated? This is where the Next Generation Scheduling comes in.

A twenty percent no-show or cancellation rate sounds very high. How do agencies currently deal with that?

There’s a lot of new technology aimed at trying to reduce the impact of that. For instance, you can call or text the riders with reminders or confirmation the night before. Mobile booking makes it easier for riders to cancel a trip in advance. For an agency, a no-show is hugely disruptive. You’ve sent a vehicle and a driver to pick a person up. There might be people waiting on the bus, and they will likely get frustrated because they’re wasting their time. A dispatcher has to react and reshuffle. One goal with Next Generation Scheduling is to be proactive rather than reactive. We will use data and business intelligence to predict with a high degree of certainty which trips may cancel. If a trip has a 90 percent chance of cancellation based on historical data or a rider’s behavior, you can start to make decisions around how you schedule the trip.

How is Next Generation Scheduling—as you say—faster, more accurate, and more efficient?

With traditional software, the scheduling engine will go through one trip at a time and slip that trip in a time slot according to the agency’s unique algorithm. If an agency has five or ten thousand trips per day, it might take between ten and thirty minutes to optimize and produce the day’s schedule. Our Next Generation Scheduling engine would do the same problem in under a minute.

Another problem for agencies is not being able to predict realistic travel times. If the schedule is just ten or twenty percent off on a 30-minute trip, those five minutes of being late will have a cascading effect on the next trips, and you can almost never recover to the original schedule times. You built a schedule that’s unrealistic to achieve. In the industry, some people call that a dispatcher’s time bomb. Our next generation engine aims to reduce or completely eliminate those dispatcher time bombs.

Why has it been difficult for agencies to predict travel times?

Traditionally, scheduling software uses map data that has a time associated with every segment of street in the map. If I want to know how long it takes to go from Point A to Point B, I add up the time on all those segments and have my answer. But what about all of the variables that can slow down or speed up a trip, like time of day and traffic patterns? That would introduce more and more data sets, which up until now, has been very difficult to consider. To do that, you need rush hours, speed zones, dozens of settings to control the time predictions. It’s complicated to setup and manage.

Instead of relying on static map data, Next Generation Scheduling is going to use the best possible source of how long it takes to move from Point A to Point B, which is the agency’s own fleet. With GPS., a vehicle is always producing a time stamp and a location as it moves throughout an agency’s service area. We can mine all that data and stitch it together so that an agency knows how long it took their vehicle to go from Point A to Point B anywhere in their street network at any time. Using this data, Next Generation Scheduling can predict the most accurate trip times in the future. Additionally, if the software recognizes that a trip is taking a lot longer than the historical models have predicted (for instance, if all the lanes are blocked on a highway), the dispatcher can be alerted, and modifications can be made for other trips.

How would this impact a rider?

With a traditional transit model, here’s how a paratransit rider might experience a trip: Say I have a doctor’s appointment at 10am. So (at least 24 hours in advance) I might ask for a pickup time at 9am. Then my trip gets negotiated to be a pickup of 8am. I arrive at my doctor’s appointment at 8:30am and wait until 10am. I finish my doctor’s appointment at 10:30 and try to book an 11am return trip, but that gets negotiated to noon, and I arrive back home at 1pm. So I had a 30-minute doctor’s appointment, and I was gone for five hours. Now, the compliance and regulatory environment will tell you that that service is fine because the transit provider delivered all the parts of that trip within the expected time window. But in reality, if you’re the consumer, it doesn’t feel inclusive and empowering. With Next Generation Scheduling, ideally the rider will save time, be more comfortable, and have the ability to be spontaneous. Combined with our Mobility and On Demand solutions, agencies can improve on that even more by offer On Demand Services.

Will Next Generation Scheduling work with multiple service providers?

Absolutely, that is the vision. Say you live in Atlanta, and you, as a rider, are eligible for three services. You can book the trip that you want to take—today or for a month from now—and we’ll figure out which provider makes the most sense to deliver that service based on algorithms and cost parameters to get you there in the timeframe we told you.

When will Next Generation Scheduling be available?

We’re launching the software now with several agencies.

What makes you excited about Next Generation Scheduling?

It addresses a lot of the fundamental problems of both our technology and the shortcomings of the other tools that are out there in a way that just makes sense. When you can give a software user a tool that is not only the most powerful and modern way of solving an industry problem but also is easy and understandable, I know the trickle-down effect will be better rides for riders. That’s an exciting thing to be able to do.

Sam Pandolfo works out of Routematch’s Atlanta headquarters as Chief Product Officer. During the college football season, he roots for the Fighting Irish at his Alma Mater, Notre Dame. Outside of the office, his hobbies are usually tied to the interests of his family, which includes his wife and two young boys. His four-year-old thinks driving a bus is much cooler than designing scheduling software for a bus company, and he frequently encourages Sam to become a bus driver. Nevertheless, Sam is proud that his work at Routematch makes an impact in the world of public transit.

Sam Pandolfo