Smart public transport and the mobility revolution

smart public transport mobility revolution

Optibus co-founder and CEO Amos Haggiag talks about the advanced smart technologies cities can use to take the lead on public transport.

When people think about the mobility revolution, they usually think about autonomous vehicles, transportation network companies providing on-demand taxi services; or micro-mobility options like scooters and electric bikes. Buses and trains have been around for a lot longer by comparison; and in many ways may seem old fashioned. However, cities have a lot more resources than many of the newer mobility options, and also have a responsibility to serve as the guardians of the public’s need for good multi-modal transportation. Public transport must seize the opportunity of a renewed public interest in mobility to become more attractive to riders both new and old.

In a world of increasing mobility options, smart public transport is increasingly necessary. But what exactly is smart public transport? In short, it is a public transportation system which incorporates advanced technology to bolster the industry’s ability to withstand the winds of change. These technologies can include the use of demand data, optimisation algorithms, distributed cloud computing and Artificial Intelligence (AI).

Demand data: where do people need to go?

Good transit is the lifeblood of any city. The ability to get where you need to go, when you need to go is essential to urban living. However, despite this, our current mobility landscape leaves much to be desired – sitting in traffic for hours is no-one’s idea of a good time.

If we want to improve mobility, we need to answer one key question: how can we take a city and move people around it in the most efficient way? One way to answer this question is to create an origin-destination matrix that considers a lot of information related to the demand for public transport, such as where people need to go, when do people need to get there, and how does demand change throughout the day. For instance, this can affect route planning, with smart transport systems using the demand data to ask about the resources that could meet people’s needs: what types of routes do we want to create, how many vehicles do we need; and how many drivers do we need?

Understanding passenger demand is an important prerequisite to creating a good transportation network. By figuring out where people are heading, transportation providers can figure out where they need to head as well, however this is only the beginning.

The proliferation of data can revolutionise how mass transport is planned today, allowing better routes and timetables to be planned to depend on variables such as day of the week, time of day and season; and therefore enabling modifications in response to real-time demand. Making fixed service less fixed may well be the next wave of innovation, revolutionising the experience of taking the bus.

Optimisation algorithms: graphing smart public transport

One reason that plotting the demand alone is not enough is that, even with unlimited goodwill, transportation providers face a variety of constraints that limit their ability to meet demand. These may include:

  • Budget constraints;
  • Number of drivers available;
  • Fleet size;
  • Depot location;
  • Environmental regulations affecting vehicle type; and
  • Labour laws and union regulations affecting crew schedule.

These limitations are not necessarily a bad thing. After all, if the budget was open-ended, operations would likely be inefficient since there would be no reason to save on cost. And if there were no labour laws or union regulations, we would be at risk of having overtired or ill-treated drivers. While regulations have important objectives, there is no doubt they function as constraints on public transport providers and greatly increase the complexity of planning and scheduling a transportation network. The fact that different providers have different constraints and preferences further adds to the complexity, making it all the more important that timetables, schedules and rosters be adapted to each individual transport provider’s needs.

The mathematical field of graph theory has yielded a method perfectly equipped to contend with complicated assortments of varied constraints and varied passenger demand: optimisation algorithms. The algorithms behind the scenes take the mess of transportation possibilities and constraints and create order from chaos, suggesting the most optimal results given the limitations. Moreover, we have found that when we tried to use readily available optimisation tools, we had to wait 24 to 48 hours for results as the algorithms struggled to work through the tens of millions of options available for each schedule.

The optimisation algorithms we have developed in-house, combined with the faster processing time of distributed cloud computing, provide results in a few seconds to a few minutes. Speedy optimisation does not just save time; it can also increase both the quality and efficiency of transport provision by opening up the user’s options. Smart public transport providers can create multiple viable schedule options, consider the different scenarios and decide which schedule works best – for cutting costs, increasing operational efficiency, improving drivers’ work benefits, and increasing the quality of the transportation network and its appeal to passengers.

Cloud computing: more flexibility, less time

In addition to demand data and optimisation algorithms, another advanced technology that can make public transport smarter and more user friendly is distributed cloud computing – the first phase in connecting the public transit network to real-time mobility-as-a-service (MaaS) options.

Cloud-native software-as-a-service (SaaS) solutions are an important element of smart public transport because they are able to capture the immense processing power of the cloud. Instead of using a single server to optimise a schedule over one to two days, a cloud-native system using distributed computing can borrow the computing power of dozens of computers for a few minutes to carry out complicated tasks requiring extensive resources. This enables optimisation algorithms to solve difficult problems in seconds to minutes instead of hours or days. This means transportation can be optimised easily even for a large city.

Schedulers using cloud-native planning and scheduling software are also able to securely access their schedule from any computer with internet access, without having to be on site, since no servers are required (and that also means they don’t have to be maintained). Someone working on a schedule can also collaborate easily with multiple schedulers or other stakeholders. In addition, cloud-native software can be used right away and is updated seamlessly, without the need for program or software update installation or IT assistance.

Artificial Intelligence: boosting on-time performance

Artificial Intelligence is another way to make public transport smarter. AI can be used to predict the on-time performance of buses by looking at how various bus routes perform in a given city, at a given time and day of the week. On-time predictions use Artificial Intelligence to learn from historical data and understand the costs and trade-offs involved – for instance, the cost implications of increasing on-time performance by a given percentage. These predictions can show the likelihood that a given bus will start or end its route on time on a particular day at a particular time.

On-time performance can also be improved with automatic scheduling suggestions which recommend how to improve the schedule to increase the overall on-time performance of the entire schedule. AI-backed on-time performance predictions and suggestions can improve operational efficiency by avoiding fines and reducing reserve levels and improve bus reliability, thus attracting more riders.

Other uses of Artificial Intelligence can include predicting ridership on dynamic routes, as well as producing dynamic timetables which enable service changes to be implemented more easily. The use of smart public transport, including advanced technologies like demand data, optimisation algorithms, cloud computing and artificial intelligence, can help public transportation providers stay competitive even as competitors continue to crowd in. The public transportation industry must not let the mobility revolution pass it by; it should take this opportunity to get into the driver’s seat.

Amos Haggiag

Co-founder and CEO


Tweet @OptibusHQ


Please enter your comment!
Please enter your name here