M2M communications, the Internet of Things, Smart Cities and now the “Industrial Internet” - every week there seems to be a new vision for how Web-connected machines will communicate and operate automatically to transform the way we live.

One of the most widely-cited predictions is that there will be more than 50 billion connected devices in the world by 2020. These devices include not only PCs, phones, and tablets but also electrical appliances such as washing machines, fridges and kettles.

Now a report by General Electric (GE) is claiming that the “Industrial Internet” has the potential to add a staggering $10-15 trillion to global GDP by 2030 and reduce billions of dollars worth of waste across major industries such as healthcare, energy and transportation.

Many companies recognise the potential of this market; chip manufacturers such as Intel and ARM have been investing in M2M processors for years, and mobile operators such as Deutsche Telekom and Telefónica have been making strategic moves in this area over the last 12 months.

Meanwhile, governments around the world are investing in “smart city” infrastructure that will purportedly enable communities to live and work in more intelligent, efficient and sustainable urban environments.

But in my view this whole issue is being approached from the wrong direction. Most of the technology needed to enable the Internet of Things already exists - sensors, networks, transmitters, receivers; the question is, what are the problems we are trying to solve?

The most over-referenced example of the Internet of Things is the fridge that sends you a text message when you run out of milk. Useful? Perhaps. Necessary? Oh please.

The only really successful implementation of M2M technology so far has been smart metering, which was born out of people's desire to be more eco-friendly and save money on their home energy bills. By monitoring the amount of energy that each domestic appliance uses, smart meters give people the visibility they need to reduce their energy consumption in a meaningful way.

But smart meters have been around for years. Why have we not moved on, and started making serious use of this technology in other sectors? Granted, there have been some small-scale projects around fleet management and healthcare, but nothing on an industrial scale.

Just imagine the amount of money and energy that could be saved by attaching a video camera and a GPS unit to the underside of every train in Britain, to monitor the condition of the tracks. The machines would be able to pinpoint and record the exact position of every fault, enabling engineers to deal with issues quickly and efficiently.

The problem is that there is still a real shortage of innovative applications for the Internet of Things. The technology is there, the big IT companies are champing at the bit to get it deployed, but the broader vision is lacking.

This, in my view, is where technology start-ups in the UK should really be focusing their efforts. Not only is there a gap in the market, but there is also a significant amount of money being pumped into the sector right now.

For example, the Technology Strategy Board (TSB) recently awarded £50,000 to 30 UK councils to complete a feasibility study showing how they could integrate their transport, communications and other infrastructure to improve the local economy, increase quality of life and reduce impact on the environment.

Larger technology companies are also looking to foster innovation in the M2M space. For example, IBM's latest SmartCamp Kickstart competition was themed around Smart Cities. The winner, Viewsy, uses smart analytics to detect signals from mobile phones so that retailers can track how many people pass through their stores, how long they are queuing and what holds their attention.

Another of the finalists, Audio Analytic, makes software that classifies sounds by computer analysis, such sounds as aggression, gunshots, glass break and car alarms. These solutions are used by leading security equipment manufactures to automatically detect and alert security operators to incidents of interest.

Public sector bodies and large firms are crying out for these kinds of companies to show them what their technology is capable of doing. And with investment opportunities on the table, the time is ripe for entrepreneurs to take full advantage.

The real challenge will be for start-up companies to think big - the Internet of Things will not be a reality until entire infrastructures can interact with one another, so it is not enough to look at solutions in isolation. They need to feed into a bigger, more ambitious strategy.

This is why start-ups need to work together with the technology giants like IBM, Intel and the mobile operators, which have an international presence and customers in various market verticals.

Software developer Living PlanIT has the right idea, with its Urban Operating System (UOS) that analyses data from a variety of sensors and automatically adjusts conditions in order to optimise efficiency.

Earlier this year, Living PlanIT signed a memorandum of understanding with Philips, McLaren Electronics, Hitachi Consulting and others, promising to work together on technologies that will enable the construction of smart cities in the UK.

If more companies can start thinking along these lines, and really take a lead on devising the applications of M2M rather than just the solutions in isolation, then the projected $15 trillion economic impact may turn out to be less fantastical than it first appears.