How to solve the problem of bus bunching

This story was first published in The Straits Times on March 22, 2013

THE recent announcements by the Ministry of Transport bode well for the public transport system in Singapore - especially for bus services.

Adding more bus lanes and enlarging bus bays will go a long way in improving service reliability. More importantly, the new bus quality incentive framework will spur the operators to run their buses on time, and standards are measured where it counts for the commuter - at every bus stop and not only at the depot.

To target buses reaching bus stops on time, the drivers will have to be empowered to make more effective decisions. But they can do so only if they are aware of their schedules at each stop; only then can they decide if they can afford to wait for that passenger running around the corner.

Many major cities such as Zurich and London have adopted a similar strategy: bus drivers have information on their schedule from onboard displays near their seat. These onboard units continuously track the position of the bus and feed it back to the traffic control centre in real time.

These units have various sensors such as GPS and odometers which ensure that the information is precise, including in tunnels and built-up urban areas.

Apart from the bus drivers, commuters too can plan their travel better if they have access to this real-time information. The result is a more effective travel system altogether.

Passengers at some bus stops can now see estimated arrival times of the next buses. The new system will increase the accuracy of this information.

The Land Transport Authority has done a good job in providing continuously updated bus schedule data to independent mobile application developers who develop apps to give commuters access to bus arrival times.

A new and more accurate system comes at a price. With the new incentives for bus operators, it might now make economic sense for them to install the system. Together with adding bus lanes and having bigger bays, this will help improve bus reliability.

But will these efforts alone ensure that the bus system is as reliable as it could be?

One critical aspect is missing from the discussion: how long a bus "dwells" at a bus stop.

Bus schedules become unreliable because of the time buses take to travel between stops, as well as the length of time each bus "dwells" at each bus stop.

Having more bus lanes and bigger bus bays clearly aims to stabilise travel times between stops. However, the issue of "dwell times" at each stop remains to be addressed.

This in fact has a direct relationship with "bus bunching", where two or more buses serving the same route arrive at the same bus stop at the same time.

How does this happen? Let's say Bus A arrives at a bus stop at 8am. There are many passengers, so Bus A spends a long time at the bus stop picking up many passengers. Bus B arrives at 8.10am, but doesn't need to spend a long time there because most passengers were picked up by Bus A.

With more passengers to drop off, Bus A's schedule is delayed, while Bus B is ahead of schedule.

At some point, Bus B is likely to catch up with the first bus, especially if the bus route is a long one. They may thus both arrive at the same bus stop at 9am midway in their route.

This causes the bus bunching that many commuters complain about, when they wait 30 minutes for a bus only to see two arriving one after the other. The "headway" between the second and the third bus increases, causing long wait times.

Given that some bus routes in Singapore are notoriously long, there is a significant potential in improving bus reliability if we adjust bus networks by splitting long bus routes into shorter ones.

New simulation studies on bus systems in Singapore conducted by the Future Cities Laboratory have shown that splitting a long bus route into two parts can potentially increase reliability by 35 per cent, as each service is required to make fewer stops. This reduces the probability of bus bunching.

Shorter bus routes also mean that buses will be more equally utilised. Fewer additional buses might be needed to serve increasing demands from a growing population, leading to significant cost savings for commuters, operators and Government.

The expansion plans for the new MRT lines announced recently will transform the public transport network. These are good reasons to think about a redesign of the bus network.

The intention however should not just be to prune or stop bus services, but to restructure bus services so they feed better into the MRT networks and serve commuters better.

For example, for longer trips, an MRT plus a short bus ride will prove a quicker commute than today's long bus rides.

Travel demand patterns for buses will change substantially when more MRT lines are ready. It is likely that there will be less need for long bus routes.

Cutting long trunk services to have more shorter bus rides may not be as popular as starting new bus routes or as visible as adding more bus lanes or enlarging bus bays. But they might prove to be efficient in improving service quality and in increasing commuter satisfaction. And that is what ultimately counts.

The writer is Research Module Coordinator, Mobility and Transport Planning at Future Cities Laboratory (FCL) which was established by ETH Zurich and Singapore's National Research Foundation (NRF).

This story was first published in The Straits Times on March 22, 2013

To subscribe to The Straits Times, please go to

Join ST's Telegram channel and get the latest breaking news delivered to you.