IBM And Chip Designer ARM Want To Monitor Every Device They Can

Cloud computing and mobile chip giants aim to master the Internet of things by crunching data from fitness gear, smart TVs, cars and more.

IBM And Chip Designer ARM Want To Monitor Every Device They Can
[Photo: Flickr user Yellowcloud]

The term “Internet of things” is a bit like “Web 2.0” was a decade ago—a buzzword for some big change in the Internet that no one can clearly describe. But as the Supreme Court once said of pornography, you know it when you see it. On a fundamental level, the IoT is about collecting data from everything—fitness bands, smoke alarms, weather sensors, cars, oil wells, even TVs and cellphones—and analyzing it to provide insights for people, companies, or marketers.


Today IBM and ARM (which licenses the designs used in billions of mobile device chips) announced an alliance to stake their claim to this new tech territory. ARM-based chips will be equipped to collect data from the devices they live in and package it securely to send to IBM’s cloud servers for processing.

What does that, and the IoT in general, mean for the rest of us?

For consumers, it could lead to gadgets that keep a closer eye on us–enabling services that deliver anything from discounts on health insurance to targeted ads on smart TVs. (Whether you want that is another matter.) For a society, it promises safer and more reliable factories, power plants, oil wells, and other major machinery, as intelligent software monitors machines and tries to anticipate breakdowns.

The latter isn’t too scary, unless you’re paranoid about Skynet. Industrial clients are already using the IoT to keep track of their machinery and plan maintenance schedules, for example, with services like IBM’s PMQ (Predictive Maintenance and Quality). Today’s deal with ARM could strengthen that process by furnishing more reliable data, if the machines use ARM processors.

IBM has a system in place to provide data analysis services for IoT devices, with its IoT Foundation processing services like Bluemix, and 44 data centers around the world.

ARM, best known for designing the brains of cellphones, says that 4 billion of its processors have been shipped for devices other than mobiles in 2014, and another billion go out every quarter. More than half of all those chips probably land in IoT devices, according to Krisztian Flautner, ARM’s general manager for IoT. “Probably. We’re not quite sure,” he says.


IBM isn’t the only company monitoring industrial machinery. GE has been offering industrial IoT services for three years to healthcare, transportation, energy and other industries. Last month, GE announced its own dedicated infrastructure, called Predix Cloud. Predix is getting started with “alpha” clients like Pitney Bowes for its package-sorting machines.

Monitoring People, Too

GE’s service is focused on big machinery, like BP’s oil wells, while IBM was already getting more personal even before today’s news. One example is SilverHook Powerboats, a maker of Nascar-like speedboats. SilverHook uploads to IBM’s Bluemix real-time data on features like fuel pressure, which, if it drops too low, can cause the engine to blow up. SilverHook also has sensors on the drivers themselves, measuring heart rate, breathing, and body temperature. “How the driver is performing and behaving affects [the race] as well,” says Bret Greenstein, IBM’s vice president of IoT.

You can expect more personal monitoring with IBM’s new IoT Service for Electronics, announced at IFA in Berlin–Europe’s biggest consumer tech show. ARM Cortex processors are in the popular Fitbit fitness trackers, for example. IBM hasn’t announced a deal with Fitbit. But Chris O’Connor, IBMs general manager for IoT, mentioned “insurance companies using Fitbit devices to change how they do premiums” as an example of what today’s deal enables. If you exercise more, for example, you might pay less–assuming you want Aetna, Blue Cross, or whoever else to know that much about you.

Expect auto insurance to figure into the mix, as well. Greenstein is a fan of Progressive’s Snapshot program (although IBM is not part of it). Snapshot uses a dongle that plugs into a car’s OBD-II data port under the dashboard and collects information such as sudden changes in speed, braking, and the times of day people drive.

Customers who appear to be safe drivers can get up to a 30% discount on their insurance. “If you offer me lower rates, I’ll provide some data,” says Greenstein.

Smart TVs will continue to wise up, too. TP Vision, which makes televisions for Philips, has signed on to IBM ‘s IoT for Electronics program. In fact, IBM is working with a number of TV makers, says Greenstein, but he isn’t authorized to name them yet.


Do people want their TVs to be so clever? “Its not really about the consumers asking for smart TVs,” Greenstein says. “It’s about engagement.” That is, companies are looking to stay in touch with customers, and ultimately make more money. If a TV maker knows what and when you watch, it can serve targeted, ad-supported TV shows and movies, for instance.

IBM says that it doesn’t have an agenda for people’s data, and that it’s providing tools to protect data. Also part of today’s announcement is the security portion of IBM’s IoT expansion. In the case of consumer products, “you want to anonymize the individual,” says O’Connor. “And we have capabilities to do that.”

Today’s IBM and ARM deal is just one of many IoT collaborations likely to pop up in the near future. As more and more Internet-connected devices fill homes and businesses, the next big question is: What do we do with all that data?

Update: An earlier version of this article incorrectly stated that ARM ships 1 billion of its processors every month, rather than every quarter. Additionally, ARM previously reported that it had shipped 6 billion processors for devices other than mobiles in 2014. The company has adjusted that figure to 4 billion, and the article has been updated to reflect that.


About the author

Sean Captain is a Bay Area technology, science, and policy journalist. @seancaptain.