Transport for London today unveiled their new bicycle hire scheme, hopefully making it easier and cheaper to get around London on short journeys. You can find details of the scheme here. The concept is pretty simple; at various locations around central London you can hire a bicycle, collecting and dropping it off at one of a number of dedicated stations, placed at strategic locations such as railway stations. Have a look at the image; it shows where the stations are located across London.
The ability to cheaply borrow a bike for short journeys is a great idea (£1 for less than 30 minute trips) and it’s been trialled in many other locations before London. However there are a few gotchas; firstly, forget to return the bike and you’ll be stung a £300 charge that covers replacement or theft. The incentive is to return it to a station as soon as you’re finished with it. Second, I’m curious as to how TFL will cope with balancing the network. There will be more demand in some locations than others for cycles; for instance in the morning, passengers arriving at railway stations will want to take a bike; at the end of the day they will want to leave one; what happens if there’s no availability of either bikes, or worse, slots to return them to? Will TFL be shipping bikes around London on the back of lorries to ensure there’s sufficient demand? If so, doesn’t that defeat the object of the scheme?
This is a perennial problem in the storage world. Storage resources are frequently unavailable in the location required; then they might not be usable because the array isn’t connected to the right fabric, or the host requires a certain vendor product. As TFL will discover, understanding and planning for demand is critical; ensuring resources are always available in the right place, without over-purchasing is a real effort, requiring planning and forward thinking.
Of course there are ways around the issues of managing demand in storage. Infrastructure design is critical; if data can’t easily be moved between storage devices (with no or minimal user intervention) then you have a problem. Designing mobility into your storage architecture is critical. Creating generic resources is too; 90% of users want storage and don’t care if it is triple-protected with a blue flashing light on the array door – just like most cyclists will be happy with a generic bicycle.
I can’t answer TFL’s problems. Perhaps they will simply rely on user supply and demand. However if you want your storage infrastructure sorted, I’m your man.