Innovative use of artificial intelligence technology is helping HSBC to ensure that our cash machines in Hong Kong don’t run out of money.

Our new iCash tool uses artificial intelligence to provide data-driven forecasting of cash withdrawals, enabling us to plan deliveries to replenish automated teller machines (ATMs) more accurately and efficiently.

Smart money (duration 0:41) Watch the animation to find out how artificial intelligence is making cash deliveries to 1,200 ATMs more efficient

Historically it has been a challenge to predict how much cash each ATM might need to meet customer demand. The process involved manually creating forecasts for demand, which could result in ATMs running out of cash or returning excess amounts, in turn causing unnecessary delivery costs.

“iCash provides a more reliable cash service to customers by making sure we have the right amount of cash in the right place,” said Chris Trill, Global Head of Wealth and Personal Banking (WPB) Operations, HSBC.

The new tool uses live ATM data and predictive machine-learning algorithms that factor in seasonality, holidays, public events, location and recent withdrawal trends to calculate how much money is needed and where.

Its dashboard visualises live withdrawal patterns, enabling the bank’s Treasury team to respond more precisely to demand and reduce lead times on cash replenishment deliveries from up to 36 hours down to just 15 minutes.

“iCash is a game-changing digital solution that improves the customer experience, while unlocking both man-hour and vendor savings,” added Mr Trill. “It also reduces the risk of robbery by moving away from scheduled cash deliveries.”

The tool, which was developed entirely in-house by HSBC’s Operations and Technology teams, has achieved a 15 per cent reduction in refill trips – saving USD1 million a year in third-party deliveries.

Further developments to iCash and a roll-out to other markets are planned.

Getting smart about using data

Ranil Boteju explores how analytics and algorithms can help banks serve customers better.

Did you find this page useful?

Why didn't you find this page useful?

Thank you. We appreciate you taking the time to give us feedback.