What have pill bottles that start screaming when pills are not taken (and even call an ambulance when no action follows, on sale in the US), the London sewers (where nobody alive knows all the pipes in existence because they remained undocumented over the centuries), and monitoring “cows’ health” have in common? They are all “things with sensors that produce data” about their status, location or behavior. Hooking up things with sensors presents boundless opportunity and implies innovative business models for smart entrepreneurs.

Rolls Royce jet engines are sold as a service, not as a product. The engines come fitted with sensors that track the status and behavior of every component. This data is then tracked in the RR Service center. When a component breaks down or needs servicing, RR notifies the necessary support people to take action or sends replacement pieces. RR adds value by turning a piece of complex machinery in a service. At present Rolls Royce gleans 51% of its revenue from servicing its engine fleet. Another example is monitoring cows’ health, by inserting them with sensors to track and monitor their condition and when they need to be milked. A tracked cow produces 200 Mb per year.

So if you can add sensors in anything, from cows to jet engines, what does that imply for other business models? We could envision embedding sensors in machines, buildings, cars, bridges, utility networks, appliances, manufacturing processes and heavy equipment to monitor them for performance, breakages, structural problems and other useful stuff. This implies that any piece of machinery and engineering can be sold “as a service” versus “as a product”. Even farm animals, geological area’s and water flows can be monitored for health and pollution levels (happens in Singapore).

Assuming this vision will become reality, who or what will manage all the underlying data and how? The traditional approach of managing IT infrastructure in house is costly. Rolls Royce does not manage their jet engine data centers on it’s own. A farmer with 500 cows will not be inclined to manage a server behind the tool shed to capture and monitor all his cow’s data (1 Gb per year and rising). The same reasoning goes for every piece of machinery, complex product or civil infrastructure that traditionally produces no data. Sooner or later they will produce massive amounts of data.

Enter cloud computing. There is authentic value in commoditizing IT infrastructure and to present various layers of IT as a service, not as hardware plus licenses. But who will offer exactly what? Will it be cloud models offered by Amazon that service the cow data management for farmers (FarmCloud)? Or will Microsoft or Google offer a one-stop-cloud-shop model which can be easily fitted to any business? Will the traditional service providers (Telco’s, internet providers) grab this opportunity and offer custom cloud services? Or will new innovative companies offer specific niche services like FarmCloud or Civil Engineering Clouds? My guess would be that it will all happen at the same time. And in true Darwinian fashion, the best adapted clouds in this new exciting environment will survive and prosper.