X

IBM: 'Internet of things' to improve infrastructure

IBM's software team is on the look out for technology to instrument and better manage basic infrastructure, such as water, energy, and traffic networks.

Martin LaMonica Former Staff writer, CNET News
Martin LaMonica is a senior writer covering green tech and cutting-edge technologies. He joined CNET in 2002 to cover enterprise IT and Web development and was previously executive editor of IT publication InfoWorld.
Martin LaMonica
3 min read

WALTHAM, Mass.--Even as much of the tech world fixates on running out of Internet addresses, Deborah Magid from IBM's venture capital group predicts many more devices not traditionally considered computers will join the Internet.

I met with Magid, who is the director of software strategy at IBM's Venture Capital group, when she was in the Boston area last week to hear about IBM's view of software start-ups and new technology. Not surprisingly, IBM is tracking software advances in mobile, cloud computing, and health care, but the energy field and technologies to improve basic infrastructure figure highly as well.

IBM is seeking to work with governments around the world to tackle societal problems on big projects, such as improving water distribution systems, reducing city traffic congestion, or making the electric grid more reliable and efficient. The computing giant, which makes more than half of its revenue from consulting-led engagements, is also involved in environmental projects, such as monitoring water quality in Galway Bay in Ireland and the Hudson River in New York.

In many of these projects, the proliferation of computing power into more and more devices, often called the "Internet of things," opens up new possibilities. "You can put technology into places that you could never do before," said Magid, who says there are already trillions of sensors in use today. "Then you have a foundation to do things that many customers don't realize is possible."

Companies, including IBM, have been talking about the Internet of things, or pervasive computing, for many years. But cheaper processors and a higher penetration of broadband networking is making that idea more of a reality, if gradually. The growth of cloud-computing services feeds the trend because embedded processors can now report data to back-end computers systems.

For example, adding more sophisticated sensors along the transmissions lines of the electricity grid can collect electricity data, such as voltage and current, and alert grid operators to potential problems before they occur.

There are more consumer-facing applications of the Internet of things, or pervasive computing, too. IBM is involved in a project in Barcelona, Spain, where parking meters convey information to alert drivers where free parking spaces are open, Magid said. Municipal leaders are pursuing the project because it improves productivity and reduces carbon pollution by 20 percent to 30 percent, she said.

For IBM, more data-producing processors create demand for back-end hardware and software in the cloud to handle the data. For the Barcelona project to work well, the system needs sophisticated algorithms based on driving patterns at different times of the day and other factors. One of the small software companies that IBM has sought to work with is U.K.-based World Sensing, which makes a software platform for writing applications that collect sensor data.

The sheer amount of data that non-computer devices produce requires a heavy dose of analytical software, too. IBM has bought a number of software companies in this field, including Cognos, DataPower, and Neteeza.

For example, a building or data center which is instrumented to use energy more efficiently by monitoring conditions in the physical setting cannot rely on people to read the data and configure heating and cooling for efficiency. Instead, the building management systems need to interface with the computing systems to make changes automatically or notify building managers of a problem.

"This is one big information management problem. A lot of it is pulling information and monitoring it. It's piles of stuff that no human could possibly understand," Magid said.