Can AI get more people to use public transport?

Artificial Intelligence (AI) can help and improve various applications, but did you ever wonder if it can improve public transport uptake? Even if public transport becomes free, similar to what happened in Tallin or Luxembourg, there’s still a big question of whether people will start using it.

Let’s be honest; if public transport is slow and unreliable, it won’t change anything if it becomes free. These problems are not due to the bus system per se but more a side effect of the congested road network and the lack of priority lanes.

Once again, AI comes to the rescue by offering solutions and insights which were before unavailable. In cities like Dubai and Beijing, they’ve installed what is known as a City Brain. These are advanced AI systems that monitor the road network operations and, through state-of-the-art algorithms, analyse massive amounts of data in real-time, thus revising the bus schedules and routes as needed.

The cities where these Brains operate reported improved public transportation and increased operational efficiency. These include a reduction in fuel consumption, a decrease in faults by 10% and an increase in ridership of almost 20%.

The reason why this happens is rather evident. We have to deal with the chicken and egg situation. Commuters don’t use the bus because the buses are inefficient. But the buses do not perform as expected because potential commuters are still using their car, thus causing congestion. In such cases, it is essential to break the vicious cycle if we want such schemes to be effective.

An AI can construct a virtual model of the road situation by analysing the bus routes and collating information from various sources (such as traffic cameras or connected vehicles). This model gives an accurate picture of what’s happening on the road and can also be used to identify meaningful patterns.

These could range from simple rules like roads within the city are less busy during the weekends than during the week. But it can also infer much more complex situations like identifying a particular route that suffers blockage during a specific time.

In this case, analysts can check the data and identify the reason. It could simply be that the path overlaps with the route of a street vendor or a garbage collection truck, thus affecting the overall efficiency. In that case, a slight diversion might solve the issue once and for all. However, the AI can also go further by automatically suggesting alternate routes, thus saving the human operator from intervening. Such an approach is faster and reduces the communication overhead humans introduce.

The AI can also analyse and improve the driving quality of the driver. Moving haphazardly across a stretch might indicate a lack of attention, eventually leading to an accident. Overspeeding in an urban area might cause a danger to others. The AI can gather these with other statistics and groom drivers by displaying friendly warnings along the route.

Every machine eventually suffers from some fault. There are various reasons for this. It could be that parts break, or extreme weather condition in hot countries causes the engine to overheat. Given enough historical data, AI systems can learn to identify patterns and create models.

There are two types of maintenance that an AI handles typically. One is called predictive, and the other is prescriptive. As the name suggests, predictive maintenance forecasts what will happen in the future.

So after analysing the data, the AI might send a particular bus for service because the vehicle will malfunction sometime throughout the day.

By doing so, the AI automatically averted an interruption of service, thus increasing the system’s overall reliability.

Prescriptive maintenance will allow the company to calculate the effects of varying the operating conditions to the time of failure. It could be that older buses are assigned less demanding routes, thus reducing their probability of developing a fault.

Given historical data, the system can also try to predict the future. If a stop is bustling during a specific period, the system can automatically summon a second bus on the same route, thus relieving the strain on the network.

Furthermore, given that the number of busses increased along a particular path, the system might instruct the driver to skip specific stops (knowing that there’s another bus on its way). Similar AI approaches in other countries have proved very effective, where they reported a reduced average waiting time of around 10% and a shorter journey time of about 5%.

Another problematic issue that public transport faces is personnel recruitment and training. Recruiting certain positions, such as technicians or engineers, might be challenging.

Because of this, they can use Augmented Reality systems to train interns or junior employees, thus ensuring that an experienced engineer aided by an AI guides them (remotely) throughout their training course. The AI will handle the most common issues, and the engineer will focus on the specific ones. Bus drivers also need extensive training. Ideally, they learn the routes and try them out beforehand.

Furthermore, they must develop strategies when dealing with particular situations, such as roadblocks. Virtual Reality (VR), together with gamification, can assist.

Through VR, they can try all the routes a priori, thus learning the path and the peculiarities along the way. The system can use a 360 film which provides high-resolution imaging and improved realism.

Apart from this, the AI will simulate dynamic obstacles along the way, thus giving them a taste of an actual situation. These simulations are essential to help them learn from experience and for management to understand the weaknesses of their staff. The training should aim to uplift the supervisory, administrative, and technical skills of inductees while familiarising them with the specialist nature of their job.

Using AI, the public transport service can provide smoother services to riders while promoting the welfare and happiness of residents. Staff will be better trained to deal with different situations, and the machines will be much more efficient.

By combining these various factors, an AI can help more people opt for public transport, similar to what it did in other countries worldwide.