Efficiently utilising data to drive improvements in the bus industry

How significant has the advance in routing and scheduling technology been in the last five years?

Rob Hudspith: We’re actually seeing some quite significant improvements and changes now. Traditionally, each element has been handled separately. You take your data, put it into Excel, analyse it, then you start your scheduling or planning in a very manual process. However, now, we’re seeing some of the software that is available being much more integrated. It takes the ticket machine data and does the analysis for you. Some systems will almost design the routes for you – they’re that good – leaving you to be able to concentrate on what you need to know in terms of designing the routes and getting the best outcome.

How much of this is technology enabled and how much is due to changes in approach at the operator level?

Hudspith: I think that it’s both. You have to adapt to the technology that’s available. Simply, why wouldn’t you want to use it to its best outcome?

Philip Gerhardt: In terms of optimising a schedule and constantly refining it, live is really key. Specifically, being able to respond live to changes in passenger flows, and potentially schedule not necessarily live, but in real time, and respond. This will be crucial in the future, because we don’t know how, or what, our passengers are going to look like. We’ve got an idea, but being able to respond quickly to ebbs and flows of demand is going to be key.

Having that tool at the fingertips of any scheduler is a fantastic opportunity. Naturally, there’ll be some who will embrace it completely and some who take a lot longer to get on board. But, I think it’s really, really important – and every part of every bus business will benefit from it.

Hudspith: Traditionally, routes are fairly fixed, frequencies are fairly fixed. As such, it takes a lot of time to make changes or to realise that something needs changing, and then push that through. If the data is there and it’s much more immediate, you can respond much quicker.

How important is it for operators and authorities to be able to scale their services to be more dynamic, especially in light of the pandemic?

Hudspith: Certainly, none of us would have ever expected to be in this situation. But, what a situation such as COVID-19 has shown is the need for at least a contingency where you can make large-scale changes very quickly. We experienced this with Transport for London (TfL) in terms of scaling down and trying to plan service levels around staffing, which is the ‘wrong way around’ from our usual way of working. The pandemic, however, has seen almost everything turned on its head. Clearly, with the drop in ridership and differing regional situations, it’s been a constant battle. In London, since beginning recovery back in summer 2020, we’ve been pretty static in terms of network service levels.

Regional businesses had to scale up and then scale back down again as the various lockdowns and restrictions came in. Importantly, it has highlighted what people can do. Two years ago, if we’d told a schedule team, ‘you’ve got to do this’, they would have thrown their hands up in horror. But, to their credit, they did it. And it teaches us a lesson for the future. There may be issues that mean we have to react quicker, and more immediate data and systems gives us the chance to make those changes more dynamically than you can.

The bus industry, in particular, always impresses me with the can-do attitude displayed from people right through the business and its schedulers and operations and engineers. They will always go the extra mile.

Thinking about the National Bus Strategy, the discussion between operation authorities is going to lead to more changes and more dynamic changes, because authorities will want to have more input, meaning that the operators, in turn, will have to react.

The network is changing in London, as is passenger demand. We’ve been looking at shifting travel patterns with TfL on their network for around three to four years now. We react to what TfL see from their data and work with them to see what we could do differently, what would give them more value for money. Conversations such as these will continue over the next few years, largely accelerated by the pandemic.

How can the wealth of operational data that’s available be best used to help operators to improve the services that they’re running? How does that data impact what comes next for operators?

Hudspith: At the moment, we use the data in terms of service reliability, identifying areas for improvements and factoring that in. Sometimes it can be too much running time, too much resource, and we can take it out where we need to, as well. But, we can be more accurate about it than just generalising, ‘this is what we think we need to do’. Going forward, the operational data, the passenger usage data is there. It’s about how that gets out and is used.

Equally, it’s external data, as well. Beyond purely what’s going through the ticket machine, looking at the information that you can get from Google on where people are visiting and times of the day and so on. We should be building that into usage, as well, and trying to predict where people are going to go. In public transport and the bus industry, we’ve always relied on what our data is telling us, rather than looking to external sources.

Gerhardt: With regards to the actual operational data, I think it’s two elements. If we rewind slightly, there’s the transparency element, and I think a lot of the operational data previously hasn’t necessarily been in the public domain. Or it has been, but there hasn’t been an interface to get it there. Now, with the open data platform from the Department for Transport (DfT), it’s all there and it’s very transparent and, consequently, it’s fair for operators and for public transport authorities (PTAs).

It’s fair for people within PTAs, or people advising PTAs, to take that data and be able to analyse it, to support bus operators in delivering the service. The key is that it’s transparent and available to everyone. And it’s not just passenger data, not just peak flows, but it’s key points of interest, it’s who is using the vehicles, times and so on.

There’s our information about what most people use, as well, and the PTAs will have that. It’s a huge way that operators can benefit from a lot of the data that PTAs are getting, too. This fits nicely into the National Bus Strategy, as well, because it goes back into a hand-in-glove partnership approach where everyone wins. It’s about unlocking doors, and I think that the mature operators, all of us, will be the ones who actually see it for what it is.

When it comes to making potential decisions about new services and new offerings, how useful is a data-driven approach?

Hudspith: If we consider some of the fixed routes where ridership is down and may not recover following the pandemic, how can the data that we have now impact decisions on rolling out more demand-responsive solutions?

When you get to that tipping point between a traditional bus service becoming unsustainable, it’s about what you do next and working between the operators and the local authorities with our historic data, plus understanding their social requirements, to see how the two things can fit together. And again, it’s taking data from external sources, not just past ticket usage, to try to predict what we can do. The more you understand why people are travelling and where they’re travelling to, you can tailor those services and generate modal shift.

It’s a vicious circle in more rural areas. Because there’s no bus service, people will drive. Because they’re driving, there’s no bus service. But, if we can predict where people are travelling or build on that data, we can, as operators and local authorities, provide something that stops them from getting in their car and delivers what they want.

If you get the data quicker, you can track more quickly how well something is working or not. And if it’s working, you build on it. If it’s not working, you could adjust it again.

Can operational data be combined with passenger data to gain further insight into new passenger requirements and improve not just existing service, but help to launch new services?

Gerhardt: Passenger requirements is such an interesting topic, because passenger requirements have changed so much. There’s a lot of companies now, in particular for trains, but also in bus, who are stating the occupancies on vehicles. I think that it’s fair to say that this wouldn’t have been a passenger requirement two years ago. An ongoing legacy of the pandemic is that passengers are using the data to avoid overcrowding and enabling them to maintain social distancing.

People are expecting more from our services, particularly in terms of hygiene – you just need to look at TfL messaging regarding cleanliness to see that. It’s health and safety – and not just about the driver being safe. It’s about the environment. It’s about the operators and, ultimately, the PTAs now, having that duty of care, and making it very clear to passengers.

Passenger requirements doesn’t just mean the journey from A to B. We talk about the experience, but the experience isn’t just, ‘Is it a nice environment?’ It’s more about, do they feel secure?

Hudspith: We’ve seen, even within our own teams, people returning to work asking if they can travel at less busy times. They have safety concerns around ‘Is the bus clean? Is the train clean? Am I going to catch COVID-19 on my journey?’

But, even when we’re past all of that, I believe that people will want to spread their days out to avoid crammed journeys. And, if you can help to provide that data, that actually helps our planning. Less of the traditional imbalance in peak travel periods makes it more efficient for us to run the full service all day or over a longer period, rather than having more buses or trains running at peak times.

Changing travel patterns and changing behaviour will affect networks. We’ve started to see that, but I don’t think that we really understand what the impact of that is going to be yet.

If our data helps people to change their travel patterns, we then need to tailor our services to make that change. Again, coming back to new services, tied with external data, which is much more available, will help us to tailor services to meet demand.