Since the launch of Deliveroo in February 2013, on-demand food delivery has experienced stratospheric growth in the UK – according to Lumina Intelligence, the market was worth £11.4 billion in 2020. Now established players such as Uber Eats, Just Eat and Deliveroo face new entrants: Estonian on-demand delivery service Bolt raised €600 million in August; two months earlier Istanbul food delivery giant Getir was valued at $7.5 billion after raising $550 million in funding.
Around five million Brits are employed in the gig economy, and while we’ve got used to the convenience they provide, scratching beneath the surface of the smooth operation of “the last mile” shows gaping inefficiencies: streets clogged by idle couriers who spend a higher percentage of their time waiting than being paid to deliver food, and who often find themselves competing for the same jobs. The companies know these issues exist, but simply factor them into their operating model.
This has not escaped the attention of academics, who are now working to apply technology to fix some of the biggest challenges of the gig economy. In 2018, a team at University College London (UCL) and Lancaster University investigated how human-computer interaction could improve efficiency and sustainability of the last-mile-delivery sector, analysing where and how delivery drivers moved around a city. At the time, they quickly found that the average speed of a delivery vehicle is just nine kilometres an hour. The reason is that the sheer number of vehicles needed to keep our e-commerce habits sated means inner-city traffic has ground to a halt. And it hasn’t improved since then. The number of vans on our roads has increased 47.5 per cent in the last 20 years, at a time when overall vehicle numbers have dropped 11 per cent. More than four million vans crawl 8.8 billion kilometres around the country every year.
In all, 62 per cent of a driver’s working day is spent with their vehicles parked at the curbside as they deliver parcels on foot. Once drivers arrived at a delivery area, they drove for around 12 kilometres, but walked for around eight km.
“We looked at opportunities for tech to improve the efficiency of logistics,” says Oliver Bates, research fellow at Lancaster University, who was behind the initial research with colleagues, including Sarah Wise and Julian Allen at UCL. Some of the researchers’ suggestions were radical: they recommended courier companies collaborate by sharing vans and drivers to try and reduce the load on city streets.
“It was a proof of concept to show how different carriers could work together, particularly in dense urban areas like London, to reduce their driven mileage and better utilise their fleets,” says Bates. One analysis by Bates and his colleagues found three couriers from three different companies spent their shifts criss-crossing each other’s routes in the same part of a city on the same day. Separate modelling by Colombian researchers found utilisation of fleets could be improved ten per cent by collaborating on deliveries, while carbon dioxide emissions could drop 25 per cent. But no one listened. During the pandemic, when the demand for home delivery increased by 29 per cent and thousands of couriers were hired by big players such as Deliveroo, drivers complained that they were spending more time waiting around and were earning less than ever before, and no delivery companies tabled the idea of joining forces.
Faced with this worsening situation, Bates and other academics have decided to not just try and fix the logistical snags of the gig economy – but to help solve the underlying problems for those engaged in this line of work, including how they’re treated by the platforms. “People who do logistics and operational research focus very much on efficiency gains when they talk about sustainability,” says Bates. They look at improving the use of vehicles and reducing the number of miles they travel, rather than considering the impact on the humans behind the steering wheel and the handlebars, he explains.