Millions of people drive from A to B every morning and then back again. Between these, their car sits empty and unused. Most people drive alone, an entire car to transport one person. This is the kind of waste that the internet and smartphones are destined to deal with - dynamic, information driven problems with real monetary effect.
So I gleefully imagined a system somewhere between a bus and a taxi; a driver picking up passengers and dropping them off, but not at predefined times and places, rather as driven by demand. No more empty buses trundling to their next vacant rendez-vous.
But most things I think up either exist or have been imagined before. So I looked for it. And in developing countries, where the costs of transport are proportionately higher and time is proportionately cheaper, there are Sharing Taxis. Or Collective Taxis. Often these will wait to fill up and then set off; the trade-off is a significantly longer expected wait for the reduction in price. I had heard of this mode of transport before, where the bus drivers will wait for a morning until they are full.
Some services act more like unscheduled buses - they are official services which travel between specific places. London apparently has some of this, although the page on the Transport for London website doesn't specify any actual information. There are also a number of community bus schemes and dial-a-ride services, largely targeted at those not covered by normal public transport; the elderly, the rurally dispersed, the disabled.
I think the time has come for a more dynamic system. One where a professional can request a service from their smartphone, indicating whether they have additional baggage or passengers, and the taxi companies can bid to take the custom. They give a price and an arrival time, and are penalised for every minute later they are. The users of the service rate their trip; largely this is a feedback mechanism for when the taxi is late, unhelpful etc.
It could begin with manual bidding and selection of services - opting for sooner and more expensive if preferred. I suspect the ultimate aim would be for taxi companies to automatically bid for each journey using live pricing systems taking in traffic data, etc. The user could subscribe to a monthly contract much like a mobile phone, which gave them a certain amount of usage. Thus the price matching would happen between the demand and supply sides.
Ultimately a market in such things, particularly one including forwards or future contracts, would allow a taxi company to buy a minibus and finance the purchase by selling a number of journeys in the near future; demand-led investment. It would also create a market for forward-looking traffic data.
The basic mechanism, however, is just that a user can request a taxi and permit sharing part or all of the journey, with a central system to connect the supply and demand sides of this; the drivers and passengers. With that in place, many daily journeys would be much cheaper, which itself enables a huge swing towards on-demand transport services (taxis).
To bring it back to the current demand-responsive transport, they could hook into such a system as providers if they already run services, or as customers if they are consumers - residents associations, charities etc. It should make both sides of that equation more economically viable as there would be both more available supply and more available demand. How? By reducing inefficiency.