RealTime IT News

IBM Stream Computing Groks Real-Time Data

IBM's new System S can help the oil industry predict the sanding of a drill bit, the medical industry foretell the onset of an illness in premature babies, astrophysicists locate the beginning of a solar flare, and banks execute global trading strategies in real time -- all in ways that were never possible before, the company said.

It's not enough to handle millions of records in a second, which IBM already does with its BAO product and SAP does with SAP Business Objects Explorer. IBM said that its Streaming Computing product is a new paradigm in which thousands of live feeds of streaming data are analyzed in real time.

"Handling data on the fly presents a unique set of challenges to the system designer. In particular, data arrives in complicated ways," Nagui Halim, IBM (NYSE: IBM) chief scientist for System S told

What IBM is calling a new science of streaming computing takes the thousands of feeds and looks only for the changes. It presents the changes in context so that decision makers can react to situations in real time.

The company cited one example in which the data stream analysis saves lives. "IBM and the University of Ontario Institute of Technology (UOIT) are using System S to help doctors detect subtle changes in the condition of critically ill premature babies. The software ingests a constant stream of biomedical data, such as heart rate and respiration, along with clinical information about the babies," the company said in a statement.

"Monitoring 'preemies' as a patient group is especially important as certain life-threatening conditions such as infection may be detected up to 24 hours in advance by observing changes in physiological data streams. The type of information that will come out of the use of System S is not available today. Currently, physicians monitoring preemies rely on a paper-based process that involves manually looking at the readings from various monitors and getting feedback from the nurses providing care," IBM added.

Autonomous learning

The system is designed to absorb expert knowledge but not require constant tending by IT. "We don't want humans handholding the model. We want autonomous learning," said Halim.

The system is designed to present data to any user, from the technologically savvy to the non-tech specialist, in the context of their work. "In manufacturing, we might show how production is going," said Halim. "We can send alerts to people. We may store data in a database for further processing later on. We can interface with mashups. For example, for pollution tracking in a watershed we would use the trajectory of weather systems as well as other data streams."

Halim added that IBM has invented a new programming language called Stream Processing Application Declarative Engine, or SPADE, because it's about digging for data.

Halim said that the language is designed to enable programmers to describe the data topology in a natural manner.

IBM is using the technology to build its Atlas for Social Connections product. Unlike other social networks, Atlas updates continuously and automatically in real time. "Participating personnel's email, chat and other streams of data are mined continuously and information in the social network is enriched and maintained current," IBM said on its Web site.

Halim said that a wide variety of customers are interested in System S including those in the healthcare, manufacturing and radio astronomy industries.

IBM System S is not yet available as a combined software plus hardware product and is not yet supported by IBM consulting services. It is a new product, announced today.