While businesses are increasingly taking to the public cloud in an effort to lower costs, Transport for London (TfL) has said it has no plans to follow suit.

In an interview with V3, TfL chief information officer Steve Townsend said moving IT services or infrastructure to the public cloud was strictly off the IT agenda for the foreseeable future.

“The public cloud is not something we’re considering,” said Townsend. “When you want the public cloud to have certain limitations, then the service becomes less flexible and there is more cost associated with it.”

Townsend suggested that complying with data regulations would mean that opting for a cloud service would not cut costs for TfL.

“When I have conversations with public cloud providers, I find that I need to know, for instance, where congestion charging data is being stored, how it is backed up and who has access to it. Often then the cost becomes prohibitive as you are essentially turning what is a public cloud back into a prohibitive private cloud,” he said.

However Townsend said TfL is making increasing use of the private cloud within its IT infrastructure. “There is definitely room for cloud-based services in the organisation. We are already using the Microsoft Azure platform for predictive services,” he noted.

“We provision train data so [outside] developers can take it and create apps that tell people when the next train is. We provision that data through the cloud so developers can easily reuse it.”

Townsend also spoke to V3 about TfL’s increasing use of big data analytics to improve transport services for Londoners. “TfL has always held the view that data is valuable but it has not always been called big data. Since the Olympic Games, we have been using big data more than ever before to look at how we can further improve business processes,” he said.

For example Townsend said analysing past data on escalator upgrades in tube and train stations may mean TfL can carry out similar projects in the future more efficiently, without the need for closing whole stations for long periods of time.

Townsend also said TfL is using big data analytics in order to run more trains per hour. “Train lines used to run 25 trains per hour as standard but now some lines, like the Jubilee and Victoria, are running 30 trains per hour,” he explained.

“Increasing the number of trains is a complex process and we have to analyse a wide range of data from dispatches information, to the number of drivers we have available on one day, to the maintenance work there is planned. Also too many trains per hour can mean the trains cannot move as fast, so this correlation also has to be analysed.

“We know we have to move more people around London on a daily basis. As we continue to upgrade, we will further use big data techniques to increase the trains per hour.”