IBM tackles M2M's big data challenge
IBM tackles M2M's big data challenge
Amid all the hoopla about machine-to-machine (M2M) applications and the vast potential of the Internet of Things to rewrite the rules of corporate sustainability, one big question tends to be overlooked.
How on earth will all that data being generated by all those devices make its way into the analytics systems and applications that can make sense of the information and turn it into actionable intelligence?
Right now, this task require racks of dedicated computer processors behind-the-scenes -- a formidable obstacle and investment for government agencies and companies evaluating the potential of M2M solutions to address everything from energy efficiency to smarter transportation systems to telemedicine.
But IBM aims to simplify the process with a new server called MessageSight. The product squeezes a process that previously required up to 280 computers into a single 2U rackmount-configured hardware "appliance" that carries an estimated price tag of less than $200,000.
This is a far more energy-efficient approach for M2M business analytics, one that IBM believes will help organizations build related solutions and services more quickly and cost-effectively than in the past.
"Big messaging is something that is an unsung hero in the middle," said Jerry Cuomo, IBM Fellow and chief technology officer for WebSphere (one of the technologies underlying MessageSight). "It gets things to work in concert with each other and broadcast what they are doing in an easy way that is easy to make sense of."
Harnessing the Internet of Things
The technology has implications for a wide range of M2M applications, particularly those in transportation and telemedicine. Wireless telecommunications company Sprint, for one, is testing IBM's new platform as a means of improving the M2M services it offers to automobile manufacturers through its Velocity telematics platform.
"To realize the vision of a Smarter Planet, we must first enable the universe of instrumented sensors, devices and machines to communicate more efficiently while sharing, managing and integrating large volumes of data at a rate much faster than ever before," said Bob Johnson, director of development for Sprint Velocity.
For example, MessageSight could help accelerate the development of services that connect the maintenance and operational sensors in cars back to the dealer, alerting them of a potentially critical issue or enabling them to unlock the doors remotely if a driver has locked himself or herself out of the vehicle.
In the future, the technology might be used to monitor the health of the person sitting behind the steering wheel, Cuomo suggested. For example, perspiration sensors could be added to the wheel grips or blood pressure monitors added to seatbelts, potentially reducing the incidence of accidents related to chronic health conditions such as epilepsy.
Alternatively, the technology could underpin in-vehicle applications that motivate drivers to adjust their habits or traveling speed to improve fuel efficiency -- or that can be used to switch hybrid vehicles between gasoline or electric modes, as appropriate automatically, without driver intervention.
An open standards approach?
The software at the heart of the solution, the Messaging Queue Telemetry Transport (MQTT), collects signals and data from M2M sensors and gets all of this to the right place for analysis to occur. The system is capable of supporting up to 1 million concurrent devices, according to IBM. Put another way, it can process up to 13 million incoming messages.
IBM is teaming up with Cisco, Tibco and other technology companies to propose MQTT as an open standard to Oasis, a non-profit trade organization instrumental in the development of Internet interoperability specifications. In short, this would enable systems to interoperate seamlessly, regardless of what sorts of sensors or devices are generating the information.
"MQTT is well suited to underpin the world of M2M/IoT and mobile applications where resources such as bandwidth and battery power are at a premium. It is an extremely lightweight, simple yet reliable protocol, designed for use where small code footprints are often advantageous," said Richard Coppen of IBM, a co-chair of the OASIS technical committee working on the standardization project.
IBM is not without potential competition in this space. Another development worth watching will be Pivotal, a venture spun out by EMC and VMware. That company just received $105 million in backing from GE, which sees Pivotal as instrumental in developing Big Data applications that could help speed adoption of its Industrial Internet initiative.