Buzzer logo

TransLink news, commentary, and behind-the-scenes stories.

Predictive departure times for real-time Next Bus

Predictive departure times for real-time Next Bus

This is one internal beta version for predictive bus departure times in map view.

In simple terms, the main question that the new TransLink mobile website has been trying to answer is: Where is my bus? Having GPS on our fleet and using the new website to track the location of buses has provided a solution to that question. Now, with the website nearing the end of its beta testing, the mobile website team is aiming to answer yet another question: When will my bus be departing from my bus stop? Providing customers with the best information to do this based on the data our systems can gather is our approach to this new and very important question. And it’s a question I hope Buzzer blog readers can help us address.

Throughout the beta phase of the new mobile site, the mobile team has been collecting feedback on the new website from Buzzer blog readers. This has been done via the comments left by readers and two in-person user testing sessions (one in August of last year and one early this month).

During this last (predictive) phase of the mobile website development, the mobile team would like to ask readers what predictive information they would like to see and how they want to see it.

How to provide predictive information

Besides data that provides the physical location of buses, which is translated into the image of a bus on the map of the mobile website, there is other information that can help customers get up to speed on the buses, routes and stops they want to know more about.

When your bus is estimated to depart from a stop

This is one internal beta version of how predictive times will look in text view.

The mobile team now has this information and is working on how best to convey it on the website. If you want to know when a bus will depart from a particular stop, all you have to do is enter the 5-digit bus stop number into the Next Bus search bar, and voila, the predicted departure time of your bus will appear.

Although Next Bus has been using Google Maps text bubbles (as shown in the example above), with predictive text, the mobile team has realized that doing so has limitations. Besides the functionality problem of when the bubble is live, the user can’t move the map. The bubble also isn’t big enough to hold enough scheduled info. Due to these drawbacks, the mobile team decided the text view was a good place to provide more detailed rider info. If you search for bus routes associated with a bus stop number in text view, you’ll see that the predictive text view allows customers to view the scheduled and predictive bus departure times. It also indicates when a bus has already passed by striking out that particular departure time. Another nice feature of predictive text view is the ability to show the bus #ID number if it can be identified by the system. If the information is not available, it will show up as “0000.″ Otherwise, it will show up in its usual four-digit ID number (e.g., 2181).

How other agencies show predictive times

Chicago’s CTA Bus Tracker mobile site shows predictive time, but no scheduled times. TriMet’s (Tri-County Metropolitan Transportation District of Oregon) TransitTracker, their mobile website, shows both predictive time and scheduled time.

TriMet's mobile site counts down to the time a bus arrives instead of using the time of day.

It could be said that showing the scheduled times when the predictive times are available is useless since what really matters is when the bus is actually going to arrive at a stop, not when it’s supposed to be there. Others might say that it’s good to have both times, so users can roughly plan their future trips.

What do you think? Would you like to see both predictive and scheduled departure times or just predictive?

Another internal beta of predictive departure times in text view. This one shows all the bus times on one line.

Next comes the important question of how the predictive text will be presented. One version of the map view of the beta version of the predictive feature (used for internal testing) indicates the exact time that a particular bus will depart a particular bus stop (e.g., 4:23 pm — see screen capture at the bottom of the post). Some systems like TriMet’s mobile website use a count down clock instead (e.g., 24 min). The advantage of showing the exact time of day that a bus will depart is that you know the time that you need to be at your bus stop to catch your bus. The drawback is that not everyone is operating on the same time. My smartphone may not show the same time as another’s smartphone, let alone someone’s wrist watch or other time keeping device. The advantage of using a count down system (as shown in these two screen captures of the beta predictive text view to the right and above) is that you’ll know you’ll need to be at the bus stop in 24 minutes no matter what time you have. The disadvantage of this system is that it requires users to do some math in order to figure out the exact time they need to be at a stop. This may be easy for departure times that are single digits, but this is more difficult and more prone to error when you need to add 53 minutes to the present time to not miss your bus.

It would be great if both the time of day and a count down to the departure of the bus could be shown, but space is limited on the map and smartphone screens, so it’s likely not an option. Perhaps a hybrid approach could be taken. Maybe a countdown could be shown if your bus is single digits away from its departure time. If your bus’s departure time is in the double digits, then perhaps the exact predictive arrival time of day could be shown. In the end, there may be drawbacks to any solution unless some smart readers can come up with another option.

Do you think predictive time should be shown as the time of day of departure, a countdown to departure or some other way?

Other data that can be provided

Some possible added information that could be included on the mobile site.

We also have access to other data including when an extra bus is added to a route, when a bus’s trip is cancelled and possibly when a bus is on a detour.

Now that we have some clear questions for Buzzer readers to weigh in on, it’s time to discuss what’s the best route for TransLink’s predictive real-time bus departure feature.

I’m genuinely elated about being able to know when I should be at the bus stop. I’m also tickled that the Buzzer blog community can help shape how it will work and look. Besides these two questions, we’re looking for feedback on all the features of real-time bus departure times. So put on your thinking headgear, readers. The future of the mobile website depends on it!


Sorry, your website browser is no longer supported.

Upgrade to one of these browsers to visit