Companies that range from General Electric to a network of hospitals in Belgium have put the Internet of Things...
concept into action to increase productivity through cloud computing and data analytics.
The Internet of Things consists of machines like electrical turbines or medical equipment outfitted with sensors that can send data to a centralized data analytics repository. New ways of processing big data can then be applied to gain new insights into manufacturing productivity or diagnostic success. All of this is well-suited for storage and processing in the cloud given the elastic nature of the data sets.
We're on the verge of a new generation of applications.
CEO, Pivotal Initiative
University Hospitals Leuven (UZ Leuven), a network of teaching hospitals located in Leuven, Belgium, is one organization that has begun to work these concepts into its IT operations.
"You can compute the genome of a human being in less than seven days," wrote Reinoud Reynders, IT manager, in an email. "One day we will have the genome sequence of all our patients and we are then in the position to compare [that] data on a regular base with reference data."
This allows clinicians to easily identify defects in the genome and can also be used to compute the chance that someone will get a type of cancer, for example, based on statistical data. This is done using software from Cartagenia Inc., which also runs on Amazon Web Services (AWS).
Not every hospital is at this stage yet.
"In our hospital we are in a unique situation in that all our medical data is in one big database, and relationships between information from different modules are very easy to do," Reynders said. "In most hospitals [that] data is separated in different systems, and then it's much more difficult to do this kind of analysis."
BlueBird Auto Rental Systems LP, based in Dover, N.J., is another company considering the Internet of Things, though it doesn't have any definite plans yet.
One thing BlueBird could do, according to Vice President Phil Jones, is have cars automatically check themselves in at each auto rental location as they are returned by customers. The problem is you don't necessarily know that every car being driven onto the lot is being returned -- the driver might want to extend the rental or just get directions. The Internet of Things would require something like a dedicated return lane at rental locations to work.
GE, Pivotal, Amazon form Internet of Things triumvirate
GE is releasing new software for its industrial customers called Proficy Historian HD, based on collaboration with the Pivotal Initiative and AWS.
Pivotal's Cloud Foundry Platform as a Service will eventually make the Hadoop-based Historian HD portable to other clouds. The Pivotal Initiative also contributed IP, which allows Historian to scale beyond GE's previous machine analytics software offerings -- meaning the software can now track years and decades of data rather than days or months. AWS will provide the infrastructure layer of the architecture for this application.
"We're on the verge of a new generation of applications," said Pivotal Initiative CEO Paul Maritz in a webcast this week. "The industrial space has to take this concept of big data to a whole new level."
Shell Oil is looking to deploy sensors inside oil wells that can generate a petabyte of data each, AWS Chief Technology Officer Werner Vogels said in the same webcast. NASA's Mars Rover also requires thousands of resources to be spun up quickly at times, and Amazon's own Amazon.com warehouses also rely on this type of technology, Vogels said. It's unclear whether any of these deployments will use the new product from GE.
The Internet of Things is not a new concept, but it's one that the industry hasn't had the technology to accomplish previously, according to David Linthicum, senior vice president for Boston-based Cloud Technology Partners.
Now, the potential exists for machines to communicate not just within, but among organizations. Tens of thousands of MRI machines, for example, could be connected to a single cloud infrastructure, and their data gathered and disseminated to people who need to make decisions, from the maintenance of the machines to medical diagnosticians.
"The sky's the limit with this stuff," Linthicum said.
That said, there are "gotchas."
"To make hay of the data and leverage it successfully is an incredibly complex problem to solve, it's an incredibly complex concept to govern, and an incredibly complex problem to secure," Linthicum said. "A lot of those question marks are still out there."