The Complexities Of Public Transport Delays

This post is also available in: Deutsch (German)

Defining public transport delays is not as straightforward as we think.

Waiting at a bus stop in Westminster Abbey
“When is that *amazing* bus coming?” said every passenger whenever there’s a delay in the bus’ arrival.
Image Source: Johen Redman

What is considered a delay?

Passengers place a very high importance on public transport punctuality. But what is considered a delay? In fact, what is the definition of delay? Let’s take the Cambridge Dictionary’s definitions of delay:

“1. To make something at a later time than originally planned or expected.

2. To cause someone or something to be slow or late.

3. To not act quickly or immediately.”

By referring to these definitions, we can assume that a public transport delay means that the transport vehicle was made to arrive slowly or at a later time than expected rather than on time or immediately.

However, passengers waiting for a bus or train may not necessarily mean that the arrival of the bus or train is delayed. For instance, if 50 passengers wait for a bus, it would feel like a big deal if the bus is delayed by 1 minute. But if no one is waiting for a bus, it won’t matter if it is late because no one is affected by the delay.

According to a study by the University of Calabria, on-time performance is among the most common measures of service reliability for public transport and can be evaluated using the percentage of transit vehicles arriving at or departing from a destination within 5 minutes (compared to the total number of arrivals and departures). This means that 5 minutes is considered as “on time”.

Meanwhile, a 2015 research paper in the UK found that passengers generally viewed a train’s arrival within 1 minute of scheduled time as being “on time”, while a short-distance train’s 10-minute delay and a long-distance train’s 20-minute delay is considered as “significantly late”.

To what extent are delays acceptable?

While there are buffers in place to define what is considered as arriving “on time”, there’s also the consideration for the level of tolerance for delays. For instance, a 2014 study in the UK found that bus passengers were tolerant about missing buses if the bus service was more frequent (every 10 minutes) but less tolerant if it’s less frequent (every 20 minutes).

In another example, a 2016 paper in the Czech Republic found that public transport passengers considered a 3-minute waiting time as very favourable, a 6-minute waiting time as favourable, a 10-minute waiting time as neutral and longer than 10 minutes as unacceptable.

The author of this study concluded that passengers who travel regularly to work already know the scheduled departures and arrive at a stop prior to a scheduled departure, so they do not expect the waiting time to increase.

It’s also proven that public transport delays could influence a passenger to book a ride via an e-hailing or a car-sharing app. A recent publication found that public transport delays increase the demand for Mobility-On-Demand services (e-hailing/taxi and car-sharing) based on data gathered in Munich, Germany.

In the event of a public transport delay, mean car-sharing demand varies by up to 13.5% while the mean taxi demand varies by up to 52.1%. If the public transport service is delayed less, the car-sharing and taxi demand can be reduced by up to 0.5% and 2.2% respectively. 

Waiting for a bus
A typical conversation at a bus stop.
Pink jacket: “This bus is so late. Wanna book a taxi?”
Red scarf: *takes out her phone from her pocket* “One step ahead of you. I already did.”
Image Source: Donald Tran

Anyone who has used public transport can relate to the problems caused by public transport delays.

Nobody wants to be late for work (unless they really hate their job) or an event they’re attending (unless they want to arrive fashionably late and irritate their aunties and uncles at their cousin’s wedding).

And nobody wants to waste their precious time on public transport because that time could have been spent in a more pleasant way, like playing sports, relaxing in the cafe or dancing on the hills.

On top of that, nobody wants to be further delayed by a train or bus that is already occupied because the delayed vehicle leads to more waiting passengers, who then board the vehicle until there is no more space.

Delays are such a pain in the neck to even think about that the list goes on and on…

What causes delays in public transport?

Until now, we’ve spoken about arrival delays but not departure delays. As the names imply, an arrival delay is when a vehicle comes late to the stop while a departure delay is when a vehicle leaves the stop late. It’s expected that a departure delay can lead to an arrival delay and vice versa.

A 2017 paper in Jordan investigating the waiting time of public transport passengers found that:

  • All public transport modes reported an overall average waiting time of 14 minutes.
  • The shortest waiting time during the peak period was about 8 minutes for a service taxi.
  • The longest waiting time was about 26 minutes for buses because of their high capacity to carry a great number of passengers and longer routes compared to other public transport modes.
  • Peak hours obviously have significant delays.

Well, it’s obvious that peak traffic conditions and long bus routes cause arrival delays for buses but there’s more to delays than just traffic jams and route lengths.

A traffic jam is definitely a factor of public transport delays (particularly bus delays) though.
Image Source: Robert Eklund

A 2009 Gershenson and Pineda paper from Mexico discussed the impact of headway instability on public transport delays. In case you didn’t know, the headway is the time interval at a station between two subsequent vehicles of the same route. We know that an equal headway is the best case scenario for any passenger since the certainty minimises the waiting times for passengers.

But that’s not always the case in reality because, the headway will be shorter if the earlier vehicle is delayed, while the headway will be longer if the later vehicle is delayed. Thus, the shorter headways lead to less passengers waiting while longer headways lead to more passengers waiting.

Why then are the headways unequal? Because vehicles move faster or slower than expected. Faster due to the tendency of some drivers to move faster than the set speed. And as we learnt in another article about public transport punctuality in addition to traffic lights and traffic jams, slower due to:

  • bus drivers (who prefer driving more slowly),
  • vehicle design and technology,
  • the lack of a bus-only lane,
  • single-train track system,
  • passenger behaviours,
  • too many passengers boarding through the middle door,
  • the combination of any of the factors listed above.

Traffic light junctions near bus stops might slow down buses while passengers who don’t move to the back seats of the buses might delay the entry of other passengers, resulting in departure delays.

Funnily, there’s even the annual issue of autumn leaves on train tracks in Ireland, causing trains to move more slowly and brake much sooner before approaching a train station.

Autumn leaves on a train track.
Autumn leaves on a train track. How much is bad enough to congest train tracks?
Image Source: Nattipat Vesvarute

Not to forget, special occasions like events and holidays can be another factor of delays or non-delays. For example, holidays can lead to a lower occupancy of public transport, thus fewer delays. Meanwhile, major events like a football match can increase the occupancy and hence, delays.

How did Covid-19 impact public transport delays?

We think that the delays have reduced. Why? The fear of contamination in public transport and the social distancing restrictions discouraging people from travelling may have contributed to less crowding in public transport.

Supporting this trend is the Deutsche Bahn, which reported that its trains were the most punctual in 2020 than the past 15 years. 81.8% of its intercity trains and 95.6% of its regional trains arrived at their stations on time (less than 6 minutes after the scheduled time by Deutsche Bahn’s standards).

However, a 2020 study by Tirachini and Cats found that public transport services during the Covid-19 pandemic have trade-offs between the need for service reliability, the need for social distancing and the need for resources.

The major drop in ridership as a result of social distancing needs and lockdown restrictions could lead public transport operators to reduce service frequency, translating to a longer headway, and thus, delays. At the same time, more resources are required to increase service reliability and enable social distancing but will incur higher costs, adding more financial strain on public transport operators.

Can we predict delays?

If you look at the psychology of waiting, you’d notice that uncertain or unexplained waits seem longer than known or explained waits. This is why it’s so important to know if a public transport vehicle will be delayed and for how long.

Thankfully, the answer to the question of whether delays can be predicted is a big YES. In addition to using Big Data to find and analyse the multiple factors of public transport delays, machine learning can be applied to predict the delays.

Combining transport schedule data and machine learning can benefit the passengers who yearn to know how much longer they have to wait for a ride.
Image Source: Sangga Rima Roman Selia

The New South Wales public transport network in Australia is using machine learning technology from Amazon Web Services to become better prepared for delays. In India, Google Maps has live traffic delays for bus travel times in India’s 10 largest cities, live train status for Indian Railway trains and multi-modal transport (combining public transport and auto-rickshaw rides) recommendations.

And as an alternative to existing and computationally more expensive methods of predicting delays in the USA, a 2020 study by Yap and Cats (yes, the same Dr. Cats mentioned earlier) proposed a supervised learning approach to predict how often different disruption types occur at different stations and their impact on delays.

The same Yap did another study in 2020 whereby he proposed a step-by-step approach to reducing the impact of disruptions:

“Step 1: Measure current disruption impacts.

Step 2: Predict future disruptions frequencies and impacts.

Step 3: Develop and evaluate measures aimed to control these disruption impacts.”

Yap (2020)

… which brings us to the final part of this article.

Dealing with delays

The issue of delays is so complex that it’s actually a very difficult task to write every point about the complexities of public transport delays. For such a complex issue, a lot needs to be done to mitigate this issue.

Perhaps, a multi-dimensional approach is needed since there are several factors of delays that can be tackled from every side of public transport services and several indicators to measure like planned arrival, real arrival, planned departure and real departure.

Among the solutions recommended for public transport operators and engineers to address every aspect of public transport delays are:

  • Holding up empty buses instead of full ones for schedule reasons.
  • Updating the passengers about delays, whether they’re happening in real-time or predicted to happen in the future.
  • Operating smaller vehicles more frequently rather than larger vehicles less frequently.
  • Paying closer attention to transfer wait times.
  • Designing methods to regulate equal headways to ensure stability.
  • Raising awareness about equal headway instability with passengers and suggesting that they take the steps listed below:

“1. If a crowded vehicle arrives at a station after a long waiting time, it is very probable that empty vehicles are coming close behind. Do not board the crowded vehicle, contributing to its further delay and of all the passengers within.

If even some people follow this advice, it is likely that crowded vehicles will be able to go relatively faster, allowing the vehicles behind them also to go faster, improving the performance of the whole system. Waiting at the station for another vehicle might actually contribute to a faster trip.

2. Give way to people descending a vehicle before boarding. Trying to “win” and enter before others will delay everybody. Sometimes waiting for a second or a third vehicle is faster than attempting to board a crowded one (especially in transport systems that allow passing).

3. Inside a crowded vehicle, go far from the doors. Giving space to ascending and descending people will accelerate the travel. Make way to the doors not too long before exiting.”

Gershenson and Pineda (2009)

Multi-dimensional Time Series Analysis VS OLAP

Public Transport Occupancy Data For Social Distancing

3 East Asian Examples Of Big Data In Public Transport

Top 5 Common Time Series Forecasting Algorithms

Categories: