An end to gridlock? How artificial intelligence could help get London moving again
London is a truly remarkable city – an economic powerhouse and a cultural destination without equal.
It’s a magnet to millions, and rightly so.
But it’s a terrible irony that this fast-moving, fast-living, fast-earning metropolis is stuck in traffic so much of the time. We simply cannot move.
According to traffic data company INRIX, London is – unsurprisingly – the UK’s most congested city. Each motorist is spending approximately 74 hours a year in gridlock during peak times.
And the effect on the environment is devastating – the mayor of London’s office concluded in 2017 that road transport in London was responsible for half of the main air pollutants in the capital.
Imagine the effect that has on the health of those living and working here. More pollution increases the rates and severity of asthma and other respiratory diseases, and also the risk of dementia – all of which add up to a massive bill for the already beleaguered NHS.
The cost of congestion is estimated at £2,430 per year per motorist – this works out at £9.5bn across the capital as a whole (up from £5.3bn in 2016).
Sadiq Khan has said that he wants 80 per cent of journeys by 2041 to be made by public transport, walking, or cycling. That would bring down the number of car journeys by three million every day – a staggering figure.
But if Transport for London (TfL) is to genuinely tackle the twin scourges of congestion and air pollution, we need to think beyond new roads and cycle highways. For long-term change, we need to look at technology, specifically artificial intelligence (AI).
To start with, AI could enable TfL to deploy buses on the routes where the most passengers are actually waiting in real-time, not just where people are expected to be.
It could also allow buses to virtually communicate with one another to ensure that they are not empty, or even half-full. Bus use would suddenly become much more efficient, and they wouldn’t travel in convoys.
If we upgraded the tech at the heart of TfL, logistical decisions could be made at the point of service delivery – by buses communicating with each other – rather than through a central hub where data is time-delayed and incomplete.
In most cities buses, taxis and trains are equipped with GPS. A good AI platform could make the most of the data that TfL already holds to improve efficiency, based on supply and demand, as never before.
Waiting passengers can already see how long it will take for their bus to get to them. Were TfL to adopt advanced technology, such as AI, the game would change completely: buses could be deployed based on the passenger and traffic volume information collected during the service.
This isn’t pie-in-the-sky thinking – this technology has recently been deployed in Barcelona by delivery startup Paack. The firm’s delivery vans and trucks were coordinated through AI and – as a result of the efficiencies made – the volume of delivery vans in the centre of the city was cut by 15 per cent. Imagine the improvement on congestion and pollution.
The United Nations has estimated that the world’s urban population will increase from 4.2bn in 2018 to 7.7bn by 2050. With this in mind, it is essential for city planners to look to alternative methods to tackle the congestion epidemic.
AI could be the fuel that TfL needs to progress along this complex and challenging journey. London just needs to give it a try.
I, for one, am sick of sitting in my car waiting for things to happen.