Data-driven solutions to the notorious 304 bus route

We’re probably all familiar with the notorious 304 bus route at this stage. I’ve been complaining about it on Twitter for months, as have many many others.

However, I’m not just a commuter, I’m a geographer, which means I can do a bit more than complain.

Last Monday 18th February, I spent a few hours taking 304 and 304A buses from Raheen to UL and back again, tracking the location of the buses over time with GPS. I collected over 13,000 data points, and analysed the results using GIS (Geographic Information Systems) software.

The data shows that there isn’t a single simple solution to the 304’s problems. In fact, analysing the data shows four different problems: dwell time at bus stops, road layout, timetabling, and route selection.

Dwell time is one of the biggest issues. The 0750 Raheen-UL 304 spent a full 25 minutes out of a 70 minute journey sitting at bus stops while passengers got on and off. That’s over a third of the trip. It’s not just a rush hour issue, too, it’s throughout the day. Even a lunchtime service spent nearly 20 minutes at stops.

Of course, that doesn’t account for all of the delays, and the bus was much slower in rush hour traffic. But it wasn’t consistently slow: there’s actually a small number of key points where the road layout holds the bus up considerably. Coming from Raheen, the bus is held up at the approach to Ballykeefe roundabout at the Crescent, and again where the bus lane ends just before Punches Cross.

Yellow is the 1520 304A, pink is the 1310 304, green is the 0750 304. Watch where the green line is held up.

Coming from UL, the approach to Groody Roundabout is a major bottleneck. Along with actually getting out of UL, those three spots add around ten minutes to a return journey.

Blue is the 0900 304A. Yellow is the 1410 304 (which skipped the city centre loop because it was 20 minutes late). Green is the 1710 304. Again, watch where it gets held up.

The surprising thing is why this isn’t reflected in the timetable, when Bus Éireann has the real time data. But the issues go even deeper than that. The 0750 from Raheen was five minutes behind timetable by the time it got to the hospital, after the initial loop through the housing estates. There was virtually no traffic to hold it up – but it spent 5 minutes picking up passengers in that time. That wasn’t unique. Basically, the only way the bus could actually keep to the scheduled timetables is to not pick up any passengers. It’s no wonder there’s constant delays, they’re literally trying to do the impossible. That’s not fair on passengers or drivers.

The good news is that there’s some easy corrections to be made which should lead to significant improvements.

First, put a second Leap card reader in buses, which passengers can use to tag on themselves. Dublin Bus has these readers just inside the doors, just tap the card on your way in, and you’re good to go. Right now, Leap cards actually add time to getting on; the driver’s machine scans for a day ticket, and when none is found the driver needs to manually enter the information – with a lot of waiting. The tap and go machines could cut dwell time in half.

Dublin Bus Leap card tag-on machine

Second, look at the routes. A 304 going to Raheen does a 15 minute loop through the housing estates, gets to the end of the line, and then turns around and does it again immediately in the opposite direction. Why? Anyone getting on during the first go around is going to be sitting on the bus as it takes half an hour to go back past where they just got picked up. Make it a loop, go around in one direction only, just like the 304A does through Monaleen after UL. That would save 15 minutes every trip.

Current (left) and proposed (right) routes for the 304 in Raheen

And why go down Henry Street and Sarsfield Street inbound and outbound? If Raheen-bound services just turned left onto O’Connell Street, that would save at least 5 minutes, more than enough time to walk from Roches Street to Debenhams.

Current (left) and proposed (right) routes for the 304 in the city centre

Make those changes, and you’re looking saving a full half an hour on every return trip – that’s around thirty hours of travel time per day. That allows you to increase the service frequency. But rather than change the timetables – just ditch them. Guarantee service frequency, and have average peak and off-peak journey times listed on the stops. There’s zero point to a timetable which is never kept to.

Longer term, I’m calling on Limerick Council to change the road layout at the Crescent and Groody roundabouts, and where the bus lane ends near Punches Cross. That would save another ten minutes, bringing the total time saved up to 40 minutes per return trip. Bus Éireann should also consider buses with middle doors when replacing rolling stock, to further reduce dwell time. But even the short term solutions would have an immediate impact, and I’m calling on Bus Éireann to introduce these measures as soon as possible.

If you would like to see more data-driven and evidence-based solutions for Limerick, I would really appreciate any support you can give for my campaign to get elected to Limerick Council in Limerick City West.

Processing ...