Bell Labs wants to turn existing fiber into a networked sensor

This story was originally published on Friday May 12, in my weekly newsletter. To receive the newsletter, sign up here. 

Sometimes the best tool to use is the one you already have. And for sensors, adding new algorithms to existing data may help provide entirely new information. That’s why Nokia’s Bell Labs division is looking at using fiber optic cables as a new type of sensor.

Mikael Mazur, a member of the technical staff at Bell Labs, said he is looking to turn the existing fiber optic cables along the sea floor, in buildings, and strung along telephone poles as a sensor to detect temperature changes, weather, earthquakes — even structural changes in buildings.

Fiber optic cable is already used as a sensor today, strung along pipelines and placed in buildings, but as Mazur explained, the features that make fiber a good sensor are the exact opposite of those that make fiber a good telecommunications material. Fiber sensors are extremely sensitive to changes in their position and vibrations, so telecommunications firms use a ton of digital signal processing to make sure the signal of the data flowing over the cable is separated from the noise of the environment around the cable.

But Mazur aims to figure out a way to use the existing fiber optic cables (and any new ones put in place for communications) as a sensor without sacrificing its ability to transmit data. For example, the subsea cables spanning the ocean could become a tool for geophysicists trying to predict the next earthquake or tsunami. Or they could provide valuable information about ocean currents or deep sea temperatures.

In buildings, fiber could be used to detect structural changes and damage caused by vibrations, while fiber laid near train tracks could detect damage to tracks and even the passage of trains.

But when it comes to sensing, the miles-long ropes of fiber currently used in telecommunications are too long. Sending data using light over long distances requires the use of optical repeaters that pick up the data as the transmission falls off and sends it along. Think of it as a relay race conducted by light and electronics, with the packets of data as the baton. Sensor readings can’t get passed along in that relay.

There are other challenges. Indeed, it’s still early days for this research. Mazur and Bell Labs have built a 524-kilometer-long (326-mile-long) aerial test cable between Gothenburg and Karlstad in Sweden to test some of the sensing capabilities of communications fiber. He presented the results of those tests in March. Research from Bell Labs may or may not make it into the real world, however, and Mazur referred my questions about timing and deployment to the optical business unit at Nokia.

If sensing does get deployed as part of a communications link, it’s unclear how sensor data would get translated into insights and be shared with interested parties. Generally communications firms care most about getting bits across a network, not what else that network can tell us.

Perhaps governmental or enterprise clients would contract with a fiber company to piggyback on the telecommunications infrastructure for their own business. It’s too soon to worry about that yet, but it’s a problem I think would be a good one to have. More sensing, especially sensing based on equipment we already have deployed, is a good thing.