How AI could help Hong Kong taxi drivers find customers on the streets
Cab payment start-up Dash and local university develop StreetSights prediction model to help taxis find high-demand areas
Hong Kong taxi drivers will be able to identify streets with high rider demand using a big data prediction model powered by artificial intelligence (AI) as soon as mid-2027, according to a cab payment start-up and a university that developed the system.
The StreetSights system is the result of a collaboration between Dash and the Hong Kong University of Science and Technology, which said on Thursday that their interim research results were up to 90 per cent accurate in forecasting demand and supply.
A taxi trade veteran said AI would be useful in securing more business as it gave drivers more data at their fingertips, noting the new regulatory regime for ride-hailing services was due to take effect in August.
Professor Lo Hong-kam, the university’s dean of engineering, said that unlike ride-hailing platforms with centrally managed fleets, street-hailing taxis faced greater uncertainty because they lacked aggregate data to give them an overview.
“There is significant room for efficiency enhancement in street-hailing taxi operations,” he said. “To do so, big data and accurate AI model calculations are absolutely essential, because there is no such flow of information [right now].”
He pointed out that the system charted offline demand and the availability of taxis across the city, allowing cabbies to decide where to go to for a higher chance of finding customers.